Assessing the Potential of Drones to Take Water Samples and

www.epa.ie

Report No.341

Assessing the Potenal of Drones to

Take Water Samples and Physico-

chemical Data from Open Lakes

Authors: Heather Lally, Ian O’Connor, Liam Broderick,

Mark Broderick, Olaf Jensen andConor Graham

ENVIRONMENTAL PROTECTION AGENCY

The Environmental Protection Agency (EPA) is responsible for

protecting and improving the environment as a valuable asset

for the people of Ireland. We are committed to protecting people

and the environment from the harmful effects of radiation and

pollution.

The work of the EPA can be

divided into three main areas:

Regulation: We implement effective regulation and environmental

compliance systems to deliver good environmental outcomes and

target those who don’t comply.

Knowledge: We provide high quality, targeted and timely

environmental data, information and assessment to inform

decision making at all levels.

Advocacy: We work with others to advocate for a clean,

productive and well protected environment and for sustainable

environmental behaviour.

Our Responsibilities

Licensing

We regulate the following activities so that they do not endanger

human health or harm the environment:

• waste facilities (e.g. landlls, incinerators, waste transfer

stations);

• large scale industrial activities (e.g. pharmaceutical, cement

manufacturing, power plants);

• intensive agriculture (e.g. pigs, poultry);

• the contained use and controlled release of Genetically

Modied Organisms (GMOs);

• sources of ionising radiation (e.g. x-ray and radiotherapy

equipment, industrial sources);

• large petrol storage facilities;

• waste water discharges;

• dumping at sea activities.

National Environmental Enforcement

• Conducting an annual programme of audits and inspections of

EPA licensed facilities.

• Overseeing local authorities’ environmental protection

responsibilities.

• Supervising the supply of drinking water by public water

suppliers.

• Working with local authorities and other agencies to tackle

environmental crime by co-ordinating a national enforcement

network, targeting offenders and overseeing remediation.

• Enforcing Regulations such as Waste Electrical and Electronic

Equipment (WEEE), Restriction of Hazardous Substances

(RoHS) and substances that deplete the ozone layer.

• Prosecuting those who out environmental law and damage the

environment.

Water Management

• Monitoring and reporting on the quality of rivers, lakes,

transitional and coastal waters of Ireland and groundwaters;

measuring water levels and river ows.

• National coordination and oversight of the Water Framework

Directive.

• Monitoring and reporting on Bathing Water Quality.

Monitoring, Analysing and Reporting on the

Environment

• Monitoring air quality and implementing the EU Clean Air for

Europe (CAFÉ) Directive.

• Independent reporting to inform decision making by national

and local government (e.g. periodic reporting on the State of

Ireland’s Environment and Indicator Reports).

Regulating Ireland’s Greenhouse Gas Emissions

• Preparing Ireland’s greenhouse gas inventories and projections.

• Implementing the Emissions Trading Directive, for over 100 of

the largest producers of carbon dioxide in Ireland.

Environmental Research and Development

• Funding environmental research to identify pressures, inform

policy and provide solutions in the areas of climate, water and

sustainability.

Strategic Environmental Assessment

• Assessing the impact of proposed plans and programmes on the

Irish environment (e.g. major development plans).

Radiological Protection

• Monitoring radiation levels, assessing exposure of people in

Ireland to ionising radiation.

• Assisting in developing national plans for emergencies arising

from nuclear accidents.

• Monitoring developments abroad relating to nuclear

installations and radiological safety.

• Providing, or overseeing the provision of, specialist radiation

protection services.

Guidance, Accessible Information and Education

• Providing advice and guidance to industry and the public on

environmental and radiological protection topics.

• Providing timely and easily accessible environmental

information to encourage public participation in environmental

decision-making (e.g. My Local Environment, Radon Maps).

• Advising Government on matters relating to radiological safety

and emergency response.

• Developing a National Hazardous Waste Management Plan to

prevent and manage hazardous waste.

Awareness Raising and Behavioural Change

• Generating greater environmental awareness and inuencing

positive behavioural change by supporting businesses,

communities and householders to become more resource

efcient.

• Promoting radon testing in homes and workplaces and

encouraging remediation where necessary.

Management and structure of the EPA

The EPA is managed by a full time Board, consisting of a Director

General and ve Directors. The work is carried out across ve

Ofces:

• Ofce of Environmental Sustainability

• Ofce of Environmental Enforcement

• Ofce of Evidence and Assessment

• Ofce of Radiation Protection and Environmental Monitoring

• Ofce of Communications and Corporate Services

The EPA is assisted by an Advisory Committee of twelve members

who meet regularly to discuss issues of concern and provide

advice to the Board.

EPA RESEARCH PROGRAMME 2014–2020

Assessing the Potential of Drones to Take

Water Samples and Physico-chemical Data

from Open Lakes

(2017-W-MS-28)

EPA Research Report

Prepared for the Environmental Protection Agency

Galway-Mayo Institute of Technology

Authors:

Heather Lally, Ian O’Connor, Liam Broderick, Mark Broderick, Olaf Jensen and

Conor Graham

ENVIRONMENTAL PROTECTION AGENCY

An Ghníomhaireacht um Chaomhnú Comhshaoil

PO Box 3000, Johnstown Castle, Co. Wexford, Ireland

Telephone: +353 53 916 0600 Fax: +353 53 916 0699

Email: [email protected] Website: www.epa.ie

September 2020

EPA RESEARCH PROGRAMME 2014–2020

Published by the Environmental Protection Agency, Ireland

ISBN: 978-1-84095-942-0

Price: Free Online version

ACKNOWLEDGEMENTS

This report is published as part of the EPA Research Programme 2014–2020. The EPA Research

Programme is a Government of Ireland initiative funded by the Department of Communications,

Climate Action and Environment. It is administered by the Environmental Protection Agency, which

has the statutory function of co-ordinating and promoting environmental research.

The authors would like to acknowledge the members of the project steering committee for their

considerate guidance and advice at various stages of the project, namely Raymond Smith (EPA), Dr

Ruth Little (EPA), Gerard Hannon (Roscommon County Council), Dr Iq Mead (Craneld University)

and Dr Cecilia Hegarty (research project manager on behalf of EPA Research).

The project team also wishes to acknowledge the valuable contributions of Mr Alan Stephens and

his team based at EPA Castlebar, research assistants Nicole Perich and Ashley Johnson, and Maura

O’Connor and Colin Folan, and Bowen Ormsby, who kindly provided boats for use on Lough Inagh

and Lough Fee, respectively.

DISCLAIMER

Although every effort has been made to ensure the accuracy of the material contained in this

publication, complete accuracy cannot be guaranteed. The Environmental Protection Agency, the

authors and the steering committee members do not accept any responsibility whatsoever for loss

or damage occasioned, or claimed to have been occasioned, in part or in full, as a consequence of

any person acting, or refraining from acting, as a result of a matter contained in this publication.

All or part of this publication may be reproduced without further permission, provided the source is

acknowledged.

This report is based on research carried out/data from March 2018 to February 2020. More recent

data may have become available since the research was completed.

The EPA Research Programme addresses the need for research in Ireland to inform policymakers

and other stakeholders on a range of questions in relation to environmental protection. These reports

are intended as contributions to the necessary debate on the protection of the environment.

iii

Project Partners

Dr Heather Lally (principal investigator,

project co-ordinator)

School of Science and Computing

Galway-Mayo Institute of Technology

Galway City

Ireland

Tel.: +353 9 174 2484

Email: heather[email protected]

Dr Conor Graham (principal investigator)

School of Science and Computing

Galway-Mayo Institute of Technology

Galway City

Ireland

Tel.: +353 9 174 2888

Email: conor[email protected]

Dr Ian O’Connor (principal investigator)

School of Science and Computing

Galway-Mayo Institute of Technology

Galway City

Ireland

Tel.: +353 9 174 2384

Email: [email protected]

Liam Broderick (principal investigator)

Model Heli Services

Belltrees

Inch

Ennis

Co. Clare

Ireland

Tel.: +353 65 683 9512

Email: [email protected]

Mark Broderick (principal investigator)

Model Heli Services

Belltrees

Inch

Ennis

Co. Clare

Ireland

Tel.: +353 65 683 9512

Email: [email protected]

Professor Olaf Jensen (principal

investigator)

Department of Marine and Coastal Sciences

School of Environmental and Biological

Sciences

Rutgers, the State University of New Jersey

New Brunswick, NJ 08901-8525

Email: [email protected]

Contents

Acknowledgements ii

Disclaimer ii

Project Partners iii

List of Figures vii

List of Tables viii

Executive Summary ix

1 Introduction 1

1.1 Background 1

1.2 Objectives 2

1.3 Project Work Packages 2

1.4 Project Dissemination 2

2 Current Use of Drones to Conduct Water Sampling in Aquatic Environments 3

2.1 Aims of the Literature Review 3

2.2 Current Use of Drones to Conduct Water Sampling in Freshwater

Environments 3

2.3 Knowledge Gaps and Technological Advances Required to Deploy Drones

for Water Sampling 6

3 Drone Platform Selection 8

3.1 Criteria for Drone Platform Selection 8

3.2 Drone Platforms 8

3.3 Costs of Drone Platform 8

4 Payload Design and Development 11

4.1 Criteria for Payload Operation 11

4.2 Key Features of the Payload Build and Design 11

5 Field Trials Deploying the Drone Platform and Payload on Open Lakes 12

5.1 Lake Sampling Sites and Experimental Design 12

5.2 Comparison of Water Chemistry Variables Collected Using Traditional Boat

and Drone Water Sampling 12

5.3 Comparison of Variability and Precision in Water Chemistry Variables

Collected Using Traditional Boat and Drone Water Sampling Methodologies 13

Assessing the Potential of Drones to Take Water Samples and Physico-chemical Data from Open Lakes

5.4 Limitations Encountered during Field Trials 15

6 Cost–BenetAnalysis 16

6.1 Efciency of Traditional Boat versus Drone Water Sampling Methods 16

6.2 Costs of Traditional Boat versus Drone Water Sampling Methods 16

6.3 Advantages and Disadvantages of Traditional Boat and Drone Water

Sampling Methodologies 18

6.4 Views Gathered from End-of-project Workshop Focus Groups 19

7 Key Findings 22

7.1 Achievements of this Research 22

8 Recommendations 23

References 24

Abbreviations 26

vii

List of Figures

Figure 1.1. Core research project work packages 2

Figure 3.1. DJI M600 Pro drone platform selected for eld trials: (a) arms and

propellers folded, (b) arms and propellers extended, (c) top view of arms

and propellers extended, and (d) drone on-site before payload attachment 9

Figure 5.1. Statistical comparison of water chemistry variable data collected using

traditional boat and drone water sampling 14

Figure 6.1. Efciency of traditional boat versus drone water sampling under different

resource scenarios 17

Figure 6.2. Benets of drone water sampling highlighted by participants during the

focus group session at the end-of-project workshop held at GMIT on

25 March 2020 21

Figure 6.3. Limitations of drone water sampling highlighted by participants during

the focus group session at the end-of-project workshop held at GMIT on

25 March 2020 21

viii

List of Tables

Table 2.1. List of research groups deploying/modifying drones to take water samples 3

Table 2.2. Specications of drone platforms used to conduct water sampling 4

Table 2.3. Specications of water sampling payloads attached to drones to conduct

water sampling 5

Table 2.4. Comparison of water sampling methods and experimental design employed

within freshwater environments 6

Table 3.1. Commercially available off-the-shelf drones suitable for adapting for water

sampling 9

Table 3.2. Costs associated with purchasing the DJI M600 Pro drone, drone platform

and accessory equipment 10

Table 5.1. Key characteristics of lakes sampled during eld trials 13

Table 5.2. Statistical assessment of precision, using coefcient of variation, between

water chemistry variable data at each sampling location between traditional

boat and drone water sampling methodologies 15

Table 5.3. List of limitations encountered during the current study when using the

drone and water sampling payload to collect water samples from Irish lakes 15

Table 6.1. Timings of activities related to traditional boat versus drone water sampling

at Ballyquirke Lough 16

Table 6.2. Capital costs associated with boat and drone water sampling during this

research project 18

Table 6.3. Advantages and disadvantages of traditional boat and drone water sampling

methodologies 19

Table 6.4. List of questions posed to participants during the focus group session at the

end-of-project workshop held at GMIT on 25 March 2020 20

Executive Summary

Water sampling remains a pivotal method for

monitoring and understanding the condition of aquatic

environments properly and effectively. Large-scale

ecological water sampling and monitoring programmes

require considerable eld personnel and are hence

resource intensive and time consuming and therefore

expensive, while also posing many health and safety

issues for personnel, as well as biosecurity risks.

Therefore, this research project had four major

objectives:

1. to assess the applicability of drones for open lake

water sampling;

2. to evaluate whether water samples and physico-

chemical data collected using drones satisfy the

European Union (EU) Water Framework Directive

(WFD) requirements for monitoring lakes in

Ireland;

3. to determine whether drones could be deployed

to increase the accuracy of extrapolated trophic

status for unmonitored lakes;

4. to examine whether drones could offer a quicker,

cost-effective, less labour intensive and safer

lake sampling protocol for use as part of the

Environmental Protection Agency WFD Lake

Monitoring Programme.

The application of drones to collect in situ

hydrochemical data and retrieve water samples from

freshwater environments provides the potential to full

some aspects of the biological and physico-chemical

sampling required for large-scale water sampling

programmes in a more efcient, safer and cost-

effective manner.

Key Findings

The project team was the rst research team in Europe

to collect a 2-L water sample using a drone. This

research has made several signicant contributions

to the advancement and application of drone water

sampling methods. These include the successful

deployment, as demonstrated during eld trials, of

a drone and attached prototype payload capable of

collecting a 2-L water sample and real-time physico-

chemical data, 100 m offshore, from lakes in the west

of Ireland. Water sampling times using the drone

were 4 minutes and water volume capture rates were

100%. Furthermore, the water chemistry of samples

collected using the drone water sampling method was

not signicantly different from that of samples taken

using a boat. In addition, accuracy and precision were

not affected by the sampling methodology employed.

This comparative analysis of water chemistry variables

satises the requirements of the EU WFD lake

water sampling and monitoring programme and thus

demonstrates that drone sampling can be applied to

large-scale water sampling programmes. The capital

investment costs for boat sampling were found to be

1.2 to 1.5 times lower than those required for drone

water sampling. However, and much more importantly,

drone water sampling was found to be 2.3 to 3.4

times faster (in person-minutes) than boat sampling

methods, depending on resource allocation. Moreover,

drone water sampling reduced both risks to personnel

health and safety and biosecurity risks associated

with boat sampling. Moreover, drone water sampling

offers a unique opportunity to sample unmonitored

lakes under the WFD in Ireland, and remote and

inaccessible lakes worldwide, and to conrm the water

quality and ecological status of aquatic environments

categorised using remote-sensing methods in a more

efcient and safer manner.

Recommendations

This research has resulted in the following

recommendations to further the advancement of drone

water sampling over the coming years:

● Consideration should be given to deploying

waterproof drones such as the Freey Alta 8 or

Alta X (https://freeysystems.com). These drones

would allow ights to operate in less than optimal

weather conditions, including in wind speeds

greater than 8 m/s and moderate rainfall.

● Smaller, lighter and cheaper real-time water

chemistry probes should be integrated to reduce

the weight of the payload and the associated

capital costs required during project set-up. This

Assessing the Potential of Drones to Take Water Samples and Physico-chemical Data from Open Lakes

would, in turn, avoid the need for larger, more

expensive drones and allow users to operate

within evolving legislative and pilot training

requirements.

● The design and build of the prototype payload,

especially its size (height and length) and weight,

should be rened to allow a greater number of

off-the-shelf drones to be considered for use.

The use of additive fabrication techniques, which

would allow the printing of new or replacement

parts when needed or damaged, should also be

considered.

● Field trials should continue and should include a

wider variety of aquatic environments including

estuaries at high and low tides, marinas

and streams and rivers at various ows, to

demonstrate application across various aquatic

environments.

1 Introduction

1.1 Background

The introduction of the European Union (EU) Water

Framework Directive (WFD) (2000/60/EC) in 2000,

and its subsequent adaptation into Irish law in 2003,

saw a Europe-wide approach to surface water and

groundwater conservation and management. The

key aim of the WFD is to ensure the good ecological

status of all European waters. For inclusion in the

Environmental Protection Agency (EPA) WFD Lake

Monitoring Programme, lakes must have a surface

area greater than 50 ha, be an active source of

drinking water or be protected under other EU

legislation such as the Habitats or Birds Directive

(Tierney et al., 2015; O’Boyle et al., 2019). To

date, 812 Irish lakes are classied as WFD water

bodies, of which a subset of 215 representative

lakes were monitored during 2013–2018 (O’Boyle

et al., 2019). For the remaining (approximately)

597 unmonitored lakes, many of which are remote

or have limited access, efforts have been made to

extrapolate ecological status using land use and

hydrogeomorphology data from monitored lakes

(Wynne and Donohue, 2016) and using macrophyte

remote-sensing data captured from Sentinel-2 data

(Free et al., 2020).

Ecological sampling and monitoring on such a large

scale requires considerable eld personnel and

is hence resource intensive, time consuming and

therefore expensive, while also posing health and

safety issues for personnel and biosecurity risks.

Currently, monitoring is undertaken by personnel from

a number of agencies including the EPA, the Marine

Institute, Inland Fisheries Ireland, Waterways Ireland,

the National Parks and Wildlife Service and local

authorities (O’Boyle et al., 2019). The WFD places

emphasis on the ecology and biology of lakes, with

physico-chemical and hydromorphological components

acting as “supporting elements” for the biota (Free et

al., 2007; Hering et al., 2010; O’Boyle et al., 2019).

The biological status of lakes is determined using

several biotic indices for phytoplankton, macroalgae,

aquatic plants, macroinvertebrates and sh (O’Boyle

et al., 2019). Physico-chemical elements that affect

the biological status of lakes include temperature,

dissolved oxygen (DO), pH, conductivity and Secchi

depth measured in the eld (Free et al., 2006).

Additional variables, such as alkalinity, colour, total

phosphorus (TP) and total ammonia, are measured

in the laboratory following the retrieval of a water

sample from the lake (EPA, 2006, 2011; Free et al.,

2006). Hydromorphological elements, which also affect

the biological status of lakes, include morphological

conditions (i.e. physical changes of the shoreline or

alterations to the natural hydrological regime) and the

water ow of the lake (EPA, 2006, 2011; O’Boyle et

al., 2019). The WFD ecological status of each lake

is assigned by applying the “one out all out” principle

whereby the lowest status achieved for any one

biological, physico-chemical or hydromorphological

element is the nal status assigned to that water body

(EPA, 2007; Tierney et al., 2015; O’Boyle et al., 2019).

The sampling of open lake waters requires the use

of a boat and this, in turn, can lead to issues related

to accessibility, particularly at remote lakes, where

there may be a lack of slipway. In 2010, 15 lakes

included in the EPA WFD Lake Monitoring Programme

were replaced because of such issues related to

accessibility (Tierney et al., 2015). Sampling using

boats can be very costly if different boat sizes are

required for monitoring different lakes and can also

lead to issues concerning personnel health and safety

and biosecurity.

Streamlining the biological and physico-chemical

sampling methods to meet the requirements of the

EPA WFD Lake Monitoring Programme could be

achieved using emerging and novel technologies.

Autonomous systems such as unmanned aerial

vehicles (UAVs), small unmanned aircraft (SUA),

unmanned aerial systems (UASs), unmanned vehicle

systems (UVSs) or remotely piloted aircraft systems

(RPASs), all commonly referred to or known as drones

(Chapman, 2014; Chabot, 2018), offer a unique

opportunity to employ novel, versatile, adaptable

and exible technologies capable of gathering high-

resolution data for monitoring and assessing the

natural environment (Wich and Koh, 2018; Fráter

et al., 2015). The application of drones to collect in

situ hydrochemical data and retrieve water samples

from freshwater environments is relatively new. The

increased capabilities of drone platforms (payload

Assessing the Potential of Drones to Take Water Samples and Physico-chemical Data from Open Lakes

weight capacity, ight time, battery endurance, etc.)

and the development of bespoke attached payloads

offers a new and unique opportunity to potentially

deploy drones in large-scale water sampling

programmes (Vergouw et al., 2016). The application of

drones provides the potential to full some aspects of

the biological and physico-chemical sampling required

for large-scale water sampling programmes in a more

efcient, safer and cost-effective manner.

1.2 Objectives

This research project had four major objectives:

1. to assess the applicability of drones for open lake

water sampling;

2. to evaluate whether water samples and physico-

chemical data collected using drones satisfy the

EU WFD requirements for monitoring lakes in

Ireland;

3. to determine whether drones could be deployed

to increase the accuracy of extrapolated trophic

status for unmonitored lakes;

4. to examine whether drones could offer a quicker,

less labour-intensive, cost-effective and safer lake

sampling protocol for use as part of the EPA WFD

Lake Monitoring Programme.

1.3 Project Work Packages

To achieve the research objectives, ve core work

packages were designed (Figure 1.1) in addition

to communications and project management work

packages.

1.4 Project Dissemination

The project team used a wide variety of approaches to

disseminate the outcomes of this project.

These included:

● one international peer-reviewed journal article;

● attendance and presentations at European and

national conferences and workshops in the eld

of drone research and water quality sampling

and monitoring [e.g. the Commercial UAV Show

2018, UK (14 and 15 November 2018); the

ShARE 5 – Shared Agencies Regulatory Evidence

Programme – meeting, Ireland (12 March 2018)

and Wales (28 January 2019); the UK and Ireland

Lakes Network Conference, Ireland (16 and 17

October 2019); and the Irish Freshwater Biologist

Association meeting, Ireland (6 March 2020)];

● the establishment of a project website and Twitter

account (@DroPLEtS18, with over 179 followers);

● the organisation and facilitation of a project

workshop attended by 21 delegates.

1. Literaure

review

2. Drone

selecon

3. Payload

build

4. Field trials

5. Cost–

beneﬁt

analysis

Figure 1.1. Core research project work packages.

2 Current Use of Drones to Conduct Water Sampling in

Aquatic Environments

2.1 Aims of the Literature Review

The aims of the literature review were to:

● evaluate the use of drones to collect water

samples and in situ physico-chemical data from

freshwater environments, synthesising and

reviewing the current literature on this topic; and

● identify knowledge gaps and technological

developments needed to advance the use of

drones to conduct water sampling in aquatic

environments in the coming decade.

2.2 Current Use of Drones to

Conduct Water Sampling in

Freshwater Environments

2.2.1 Research teams deploying/modifying

drones to conduct water sampling

The literature review highlighted several research

teams, predominantly in the USA, Japan and Australia,

working on deploying/modifying drones to take water

samples (Table 2.1).

2.2.2 Specicationsofdroneplatformsused

to conduct water sampling

Specications of drone platforms used to conduct

water sampling varied among the various research

groups. Over the past decade, a combination of off-

the-shelf drones (e.g. Ascending Technologies Firey

hexarotor, six rotor LAB645 UAV and DJI Matrice 600)

(Detweiler et al., 2015; Ore et al., 2013, 2015; Song et

al., 2017; Castendyk et al., 2018, 2019, 2020; Terada

et al., 2018) and custom-built platforms (Koparan

and Koc, 2016; Koparan et al., 2018a,b, 2019) have

been deployed (Table 2.2). Drone platforms have

a maximum payload weight of between 600 g and

12 kg and a maximum ight time of 20–40 minutes

depending on the amount of water to be collected.

Some research groups apply autonomous operating

systems and/or pilot-operated systems.

2.2.3 Specicationsofpayloadsusedto

conduct water sampling

All water sampling payloads are custom-built with

various types of water sampling systems having

been trialled, including (1) complex chassis systems

with three spring-lidded chambers operated by a

servo-rotated “needle” where water lls a glass

sampling container via a micro submersible water

pump (Detweiler et al., 2015; Ore et al., 2013, 2015),

(2) triple cartridge “thief style” water sampling systems

(Koparan et al., 2018a, 2019, 2020), (3) Niskin water

sampling bottles (Castendyk et al., 2018), (4) high-

density polyethylene (HDPE) bottles consisting of a

hollow tube structure that allows water to freely enter

when lowered into the water (Terada et al., 2018), and

(5) HydraSleeve (Castendyk et al., 2020) (Table 2.3).

Trials have demonstrated the ability of the water

sampling payloads listed above to collect between

Table 2.1. List of research groups deploying/modifying drones to take water samples

Name of research lab Location Publications

NIMBUS research lab Nebraska-Lincoln, USA Scientic publication(s)

Chung research lab UC Berkeley, USA Scientic publication(s)

Koparan research lab Clemson University, South Carolina, USA Scientic publication(s)

Terada research group Japan Scientic publication(s)

HATCH Associates Consultants Denver, Colorado, USA Scientic publication(s), conference publication

Rise Above Custom Drone Solutions Australia Website information

Spheres Drones Australia Website information, leaet

UC, University of California.

Assessing the Potential of Drones to Take Water Samples and Physico-chemical Data from Open Lakes

60 mL (Detweiler et al., 2015; Ore et al., 2013, 2015)

and 2 L of water (Castendyk et al., 2019) with water

sampling times ranging from 40 minutes (Song et

al., 2017) to 2 hours (Detweiler et al., 2015; Ore

et al., 2013, 2015) and successful water capture

rates varying between 60% (Ore et al., 2013, 2015;

Koparan and Koc, 2016; Koparan et al., 2018a)

and 100% (Koparan et al., 2019). Issues with the

complex chassis systems with three spring-lidded

chambers were predominantly associated with faulty

lid mechanisms, variations in the altitude of the pump,

the pump not priming correctly, silt intake in the pump

and environmental conditions such as increases in

wind speed above 2.7 m/s, which signicantly reduced

sample capture success, with a linear decline with

increasing wind speed resulting in a 25% decrease

in sample capture at speeds in excess of 4.5 m/s

(Ore et al., 2013, 2015). Issues associated with

the messenger on the “thief style” water sampler

included the sampler not triggering or the servo-motor

malfunctioning (Koparan and Koc, 2016; Koparan et

al., 2018a).

Research teams have also incorporated off-the-shelf

multi-meter probes (temperature, DO, conductivity and

pH) (Song et al., 2017; Koparan et al., 2018b, 2019),

conductivity, temperature and depth (CTD) probes

(Castendyk et al., 2018, 2019, 2020) and turbidity

(Koparan et al., 2020) with the ability to autonomously

relay real-time data back to nearby ground stations

(Song et al., 2017), greatly increasing the capacity of

drones to monitor in situ water chemistry variables.

2.2.4 Comparison of water chemistry

variables

Comparison of water chemical variables obtained

using drone water sampling with those of more

traditional methods (e.g. handheld probes and

manual grab samples from land or boat) shows clear

Table 2.2. Specications of drone platforms used to conduct water sampling

Platform type

Maximum

payload

weight Flight time Communication software Source

Off-the-shelf

hexarotor –

Ascending

Technologies

Firey

600 g

Total ight

time = 15–20

minutes per

battery with full

payload

• Robot operating system – low-level communication

with the UAV, risk management, mission control,

navigation and altitude estimates

• Onboard custom microcontroller – operates aerial

sampling system, reads sensors and status of water

sampling system

Detweiler et al.,

2015; Ore et

al., 2013, 2015;

Song et al., 2017

Off-the-shelf six-

rotor LAB645

12 kg

Maximum

ight time = 40

minutes

• Operator controlled during take-off and landing

• Autonomous ight via GPS waypoints

Terada et al.,

2018

Off-the-shelf DJI

Matrice 600 Pro

6 kg

Maximum

ight time = 16

minutes

• Operator controlled during take-off, ight mission and

landing

• Spotter ensuring “safe ight area”

Castendyk et al.,

2018, 2019, 2020

Custom-built

hexacopter

with otation

attachments

750 g

Theoretical ight

time = 8 minutes

• Radio controller (Turnigy 9X) – manual control of

hexacopter

• Autonomous ight and ground control station –

Pixhawk autopilot (GPS receiver, radio telemetry) –

provides information on ight conditions

• Mission planner software

Koparan and

Koc, 2016

Koparan et al.,

2018a,b

Custom-built

hexacopter

with otation

attachments

2.1 kg

Theoretical ight

time = 8 minutes

• Radio controller (Turnigy 9X) – manual control of

hexacopter

• Autonomous ight and ground control station –

Pixhawk autopilot (GPS receiver, radio telemetry) –

provides information on ight conditions

• Mission planner software

Koparan et al.,

2019

Manufacturers’ specication.

Weight of components used to develop the custom-built payload.

Flight time based on 80% battery life.

GPS, global positioning system.

H. Lally et al. (2017-W-MS-28)

inaccuracies (Table 2.4). These can be attributed

to variations in water sampling payloads used, the

manner in which drones were deployed to collect

water samples and physico-chemical data, and limited

experimental design.

Ore et al. (2013, 2015) and Detweiler et al. (2015)

reported similar trends for physico-chemical variables

collected using drone and manual water sampling

methods. However, DO levels were higher and

temperature levels lower in waters collected using

Table 2.3. Specications of water sampling payloads attached to drones to conduct water sampling

Sampling

location

Water sampling

payload

Physico-

chemical

sensors

attached to

drone

Quantity

of water

collected

Water sampling

times using

drone

Physico-chemical

variables monitored Source

Holmes Lake

(Nebraska,

USA)

Custom-built

chassis – spring-

lidded chambers

operated by a

servo-rotted

“needle” with tube

and micro pump

None 60 mL Total time =

2 hours

Estimate

20 minutes using

the drone alone

Temperature, DO,

sulphate and chloride

Detweiler et

al., 2015; Ore

et al., 2013,

2015

Mesocosms,

University

of Kansas

Biological

Field Station

(Kansas, USA)

As above Temperature

(GP103J4F NTC

Thermistor) and

conductivity

(Atlas Scientic)

sensors

As above Total time =

40 minutes

10 minutes per

reading per

mesocosm

Temperature,

conductivity and

chloride

Song et al.,

2017

Yugama crater

lake (Japan)

Custom-built

metal-free HDPE

sampling bottle

None 250–330 mL Not given Conductivity, pH,

chemical concentration

(chloride, sulphate,

aluminium, calcium,

iron, potassium,

magnesium,

manganese, sodium,

silicon dioxide) and

stable isotope ratios

(δD and δ

Terada et al.,

2018

Lamaster

Pond, Clemson

University

(South

Carolina, USA)

Custom-built

triple-cartridge

“thief style” water

sampler

pH, conductivity,

temperature

and DO (Atlas

Scientic)

sensors

130 mL Total time = 1 hour

Estimate

20 minutes using

the drone

DO, temperature,

pH, conductivity and

chloride

Koparan and

Koc, 2016;

Koparan et

al., 2018a,b,

2019

Lake

Issaqueena

(South

Carolina, USA)

Custom-built

triple-cartridge

“thief style” water

sampler

Turbidity sensor

(DFRobot)

130 mL Total time = 1 hour

Estimate

20 minutes using

the drone

DO, temperature,

pH, conductivity and

chloride

Koparan et

al., 2020

Pit lakes

(Ontario,

Canada, and

Nevada, USA)

Niskin water

bottle (General

Oceanographics,

Florida, USA)

Conductivity,

temperature and

depth (CTD)

(YSI CastAway)

probe

1.2 L Flight time of

drone is less than

15 minutes

None Castendyk et

al., 2018

Pit lakes

(Montana and

Idaho, USA)

As above As above 2 L As above pH, calcium,

magnesium, sodium,

chloride, sulphate, total

dissolved solids, total

organic carbon, total

sulphide, potassium

Castendyk et

al., 2019

Pit lakes

(Nevada,

Montana and

Idaho, USA)

HydraSleeve

(GeoInsight)

As above 1.75 L As above Temperature, specic

conductance,

bicarbonate alkalinity,

chloride, sulphate,

calcium, potassium,

sodium, cadmium,

manganese and zinc

Castendyk et

al., 2020

Assessing the Potential of Drones to Take Water Samples and Physico-chemical Data from Open Lakes

the drone water sampling method. Differences were

attributed to interference or contamination of carryover

by the pump and transit through the tubing, agitation

during ight and in some instances changes in water

properties between water collection and analyses

(Detweiler et al., 2015; Ore et al., 2013, 2015). In

comparison, sulphate and, in particular, chloride

levels were lower in samples collected using the

drone water sampling method than in simultaneously

collected manual grab water samples. Differences

were pronounced; values were deemed to be due to

sampling variation (Detweiler et al., 2015; Ore et al.,

2013, 2015).

Comparative statistical studies by Koparan et al.

(2018a) found drone water samples to be signicantly

higher for DO, pH and chloride. However, the

percentage difference between water chemistry

variables was deemed small, highlighting minimal error

between the sampling methods. Song et al. (2017)

also reported signicant differences in the levels

of chloride between drone and manual grab water

samples, which were attributed to interferences within

the water column from using a boat and differences

in the volume of water collected. The small volume

of water collected (20 mL) using the drone sampling

method may have been less representative of the

chloride levels.

Finally, Castendyk et al. (2020) found ex situ handheld

multi-parameter probe readings were 2.8 degrees

higher for temperature than in situ drone water

samples using CTD, although this temperature

increase was attributed to natural warming during

water retrieval, presumably due to the differences in

air and water temperatures and the lack of insulation

on the sampling device. In addition, they reported

contract-required detection limits (CRDLs) exceeding

ve and relative percentage differences (RPDs) of less

than 20% for a wide suite of water chemistry variables

(Table 2.3) with the exception of cadmium and

manganese. However, all samples were deemed to

be within the range of the CRDL and therefore drone

sampling methods were considered equivalent to

those collected using traditional boat sampling under

US EPA guidelines (US EPA, 1994).

Overall, limited sample size and replication and poor

water capture rates may have prevented robust

statistical comparison between water sampling

methodologies in many of the reviewed studies

(Tables 2.3 and 2.4).

2.3 Knowledge Gaps and

Technological Advances Required

to Deploy Drones for Water

Sampling

2.3.1 Current knowledge gaps to deploying

drones for water sampling

The studies reviewed highlight the potential use

of drones to conduct water sampling and obtain

physico-chemical data from freshwater environments.

However, several key limitations were highlighted,

which need to be addressed before the application of

Table 2.4. Comparison of water sampling methods and experimental design employed within freshwater

environments

Sources

No. of

sampling

sites Replication Methods compared

Total sample

size

Statistical

comparison

Ore et al., 2013,

2015; Detweiler et

al., 2015

5 3 Manual grab sample and use of handheld

probes from kayak vs drone-assisted water

sampling from kayak

30 None

Song et al., 2017 9 3 Manual grab sample and use of handheld

probes vs HOBO in situ sensors vs drone-

assisted water sampling

81 None

Koparan et al., 2018a 3 3 Manual grab samples and use of handheld

probes from kayak vs drone-assisted water

sampling

18 Paired t-tests

Castendyk et al.,

2020

1 3 Manual van Dorn samples and use of

handheld multi-parameter probes from boat

vs drone-assisted water sampling

6 RPD

RPD, relative percentage difference.

H. Lally et al. (2017-W-MS-28)

drone technology can be applied to large-scale aquatic

sampling programmes worldwide.

Consideration should be given to the following:

● type and payload capacity of off-the-shelf drones,

as many new large drones (less than 25 kg) have

a payload carrying capacity of at least 10 kg (see

Chapter 3, Table 3.1) and could be modied to

allow larger volumes of water (minimum 1–2 L) to

be collected;

● ability of the drone to successfully complete ights

in less than optimal weather conditions if larger

volumes of water are to be obtained;

● improving sampling success rates, if larger

volumes of water are to be captured every time;

● obtaining samples beyond visual line of sight

(BVLOS) (greater than 300 m) (EASA, 2015,

2018; ICAO, 2011; JARUS, 2013) especially on

large open water bodies (greater than 2 ha or

20,000 m

);

● increased costs, set-up times and legal

requirements of deploying larger drones and

associated payloads, and BVLOS.

2.3.2 Recommended technological

advancements required to deploy drones

for water sampling

Technological advancements in water sampling

payload design are required if accurate and reliable

statistical comparisons of water chemistry variables

are to be determined.

Consideration should be given to the following:

● incorporating in situ, real-time data of physico-

chemical parameters (temperature, DO,

conductivity and pH) for meaningful comparisons

of data obtained using drone and handheld

sampling methods;

● deploying the same probes (from the same

manufacturer) when comparing drone and

handheld water chemistry data for clearer,

transparent and consistent comparisons that

are not potentially confounded by differences

in performance due to different types of probe

models;

● allowing sampling probes time to take readings

(suggested minimum time of 4 minutes) when in

the water, with sampling times dependent on both

the probe used and physico-chemical conditions

of the water body being sampled; for example, the

minimum time is usually determined by the pH of

the water body, with lakes with low pH and low

buffering capacity generally requiring longer times

than calcium-rich lakes;

● adapting robust statistical experimental designs

to examine the data collected between sampling

methodologies as well as the variability;

● incorporating a greater number and diversity

of types of water bodies, an increased number

of sampling sites per water body and greater

replication of samples per sampling site per water

body;

● testing a wider selection of water chemistry

parameters (nutrients, suspended solids and

heavy metals) and comparisons across various

water sampling methodologies (manual grab

samples and handheld probes vs drone water

sampling vs in situ sensors);

● conducting a detailed cost–benet analysis.

This literature review, entitled “Can drones be used

to conduct water sampling in aquatic environments?

A review”, by Lally, O’Connor, Jensen and Graham

is published in Science of the Total Environment,

670 (2019), 569–575 (reused with permission from

Elsevier).

3 Drone Platform Selection

3.1 Criteria for Drone Platform

Selection

The research team developed a set of criteria specic

to the needs of the proposed project in addition to

setting out the specications of the drone platform

required.

The criteria and specications included:

● the requirement for a drone with a stabilised

three-axis zoom camera, foldable rotor arms and

propellers;

● a drone platform able to lift a minimum slung

payload of 6 kg;

● the ability to y for a minimum of 10 minutes and a

distance of 200 m carrying a 6-kg payload weight;

● the capacity to y in wind speeds up to 8 m/s;

● the capability to provide “loss of a single motor”

redundancy;

● having global positioning systems (GPS) and

global navigation satellite system capabilities with

redundancy between them;

● having a retractable undercarriage that does not

interfere with the payload;

● having a rst person view integrated zoom camera

of at least 12 megapixels;

● being suitable for dual operators where the

camera movements and control can be controlled

by the second operator with two compatible pilot/

camera high-denition monitors supplied for live

view with an HDMI (high-denition multimedia

interface) output;

● the capability for the primary controller to provide

additional channels for integration with the

payload;

● having a controller system that allows for software

development kit integration;

● having radios that operate on primary EU

industrial, scientic and medical (ISM) band

2.4 GHz transmission with optional 5.8 GHz

backup transmission;

● a platform that gives real-time telemetry

information for the duration of the ight;

● a platform and all its accessories that are

CE approved.

3.2 Drone Platforms

Eight off-the-shelf drone platforms were investigated

based on the criteria set out in section 3.1 above

(Table 3.1). A DJI Matrice 600 Pro platform [Shenzhen

Dà-Jiāng Innovations (DJI) Sciences and Technologies

Ltd, Shenzhen, Guangdong, China], hereafter referred

to as the DJI M600 Pro drone, best matched the

criteria set out by the project team and was therefore

selected for use in eld trials (Figure 3.1). The

DJI M600 Pro drone was registered with the Irish

Aviation Authority (IAA) through its ASSET drone

registry programme operated by CGH Technologies

Incorporated. Within the Marine and Freshwater

Research Centre (MFRC) at the Galway-Mayo

Institute of Technology (GMIT), the project team

updated the MFRC Drone Operations Manual and

risk assessments to include the DJI M600 Pro drone.

In addition, the platform was added to the GMIT’s

insurance policy.

3.3 Costs of Drone Platform

The unit cost of the DJI M600 Pro drone was

€11,370.27 including the costs of purchasing

the drone platform and accessory equipment

necessary to modify the undercarriage and allow for

communications with the water sampling payload

(Table 3.2).

H. Lally et al. (2017-W-MS-28)

Table 3.1. Commercially available off-the-shelf drones suitable for adapting for water sampling

Name of

UAV Manufacturer

Size

(diagonal)

(mm)

UAV

weight

(kg)

Maximum

payload

weight

(kg)

Flight

speed

(m/s)

Flight

duration

(minutes)

Wind

speed

(m/s) Waterproof

EU regulatory

legislation

Matrice

600 Pro

DJI

1133 10 6 17 16 8 Splashproof Open category – A3

operator with IAA

Conduct theory and

ight exams

Hold third-party

insurance

Alta 8 Freey

1325 6.2 9 – 17–20 – Weather

resistant

Agras MG-

1P and

MG-1S

DJI

1500 and

1515

10–14

7 20–22 8 Sprayproof

Alta X Freey

1415 10.4 15.9 – 10 – – Specic category –

SOP required

operator with IAA

Include risk

assessment

Skymatrix

X-FI

Prodrone

1534 13.2 20

16 13–25 8 Water- and

all weather-

proof

Agras T16 DJI

1833 18.5

16 10 10 8 Sprayproof

PD6B –

Type II

Prodrone

1348 11.5 30 16 10–30 10 –

Data on DJI UAV models were taken from the DJI ofcial website (www.dji.com).

Data on Freey models were taken from the Freey website (www.freeysystems.com).

Data on Prodrone models were taken from the Prodrone website (www.prodrone.com).

Weight excludes batteries.

Payload is a spray tank.

EU regulatory legislation pertains to the new EU Implementing Regulation [Commission Implementing Regulation (EU)

2019/947] and Delegated Regulation [Commission Delegated Regulation (EU) 2019/945], which took effect on 1 July 2020. See

www.iaa.ie/general-aviation/drones for more information.

SOP, special operating permission.

Figure 3.1. DJI M600 Pro drone platform selected for eld trials: (a) arms and propellers folded, (b) arms

and propellers extended, (c) top view of arms and propellers extended, and (d) drone on-site before

payload attachment (photo credits: Heather Lally).

(a) (b)

Assessing the Potential of Drones to Take Water Samples and Physico-chemical Data from Open Lakes

Table 3.2. Costs associated with purchasing the DJI M600 Pro drone, drone platform and accessory

equipment

Description Unit price excluding VAT (€)

Drone platform

DJI M600 Pro drone 5522.70

DJI CrystalSky monitor 7.85 ultra brightness 1137.75

DJI CrystalSky bracket 90.00

DJI M600 propellers (one CW set/one CCW set) 61.50

Total 6811.95

Batteries

DJI M600 ight pack (four sets of six TB47S batteries) 1102.08

DJI M600 Hex charger UK version including UK plug 332.10

DJI M600 battery case 584.25

Total 2018.43

Undercarriage accessories

DJI Z3 camera 956.94

DJI M600 camera mounting plate for Z3 camera 209.10

Hooking system 130.00

DJI power hub 20.00

Total 1316.04

Communication accessories

DJI M600 remote controller 774.90

DJI M600 channel expansion module (eight channels) 448.95

Total 1223.85

Total unit cost of DJI M600 Pro 11,370.27

CCW, counter-clockwise; CW, clockwise; VAT, value added tax.

4 Payload Design and Development

4.1 Criteria for Payload Operation

The research team developed criteria for the operation

of the payload specic to the needs of the proposed

project.

The criteria included the following:

● It must be able to retrieve 2 L of lake water from

approximately 10 cm beneath the surface.

● It must be easily decontaminated to prevent the

potential spread of invasive species.

● The design must include the ability to ush the

system after each sample/site/lake to ensure no

contamination of subsequent samples, including

chemicals associated with biosecurity.

● The water sample must not be contaminated by

any materials (e.g. metal components) used to

develop the payload.

● It must be possible to collect and log “time

stamped” chemical measurements using a data

logger when the payload is in the water.

● The nal payload weight should be suitable for

transportation using the DJI M600 Pro drone.

● Payload operation from water entry to exit must be

automated and triggered from the drone console.

● The payload should be waterproof and completely

buoyant.

The payload design and build are proprietary and not

described herein.

4.2 Key Features of the Payload Build

and Design

Key features of the payload build and design include

the following:

● The water sampling bottles used are 1 L, HDPE,

opaque and wide mouthed.

● Real-time physico-chemical data are transmitted

via a YSI EXO Go and EXO Sonde with EXO

pH, DO, temperature and conductivity probes.

The EXO Go communicates with the Sonde

using Bluetooth and when paired with a Windows

operating system device with KorEXO software

the data can be live streamed in real time to the

user on the shore.

● The prototype payload weight is 6 kg outbound

and 8 kg inbound.

● The water sample weighs 2 kg.

● Samples are taken more than 100 m from the lake

shore.

● The ight time from take-off to return is 6 minutes,

including a water sampling time of around

4 minutes.

● The altitude for the ight is maintained at less than

15 m in all cases.

● The collection of a 2-L water sample and in situ

real-time chemical analysis are performed with

100% reliability.

● A remotely operated camera and “live link” verify

payload operations.

● Operational testing on site veries that safe “one

person” take-off and landing remote from the

shoreline can be deployed in difcult locations if

required.

● An emergency release has been incorporated and

tested and can be deployed should the payload

become fouled or entangled or drone performance

deteriorate.

5 Field Trials Deploying the Drone Platform and Payload

on Open Lakes

5.1 Lake Sampling Sites and

Experimental Design

Field trials took place between September and

November 2019 at six lakes (Loughs Fee, Inagh,

Conn, Derg and Mask and Ballyquirke Lough) in the

west of Ireland. The lakes chosen represent two of

the main lake types found in Ireland (high and low

alkalinity) and a range of trophic gradients. Loughs

Fee and Inagh are currently not monitored under

the EPA WFD Lake Monitoring Programme, while

Loughs Conn, Mask and Derg and Ballyquirke Lough

are included in the monitoring programme. Key lake

characteristics are presented in Table 5.1.

One location was sampled on Loughs Fee and Inagh,

two locations on Lough Conn and Ballyquirke Lough,

and three locations on Loughs Mask and Derg. At

each location, three paired water samples (traditional

boat water sampling and drone water sampling)

were collected (n = 36 paired water samples in total).

Data and samples were collected at each location to

examine the variability associated with each sampling

method. On completion of sampling, water samples

were delivered to the EPA laboratories in Castlebar

for analysis of pH, alkalinity (mg/L CaCO

), hardness

(mg/L CaCO

), true colour (mg/L PtCo), chloride

(mg/L), silica (mg/L SiO

), ammonia (mg/L N), total

oxidised nitrogen (TON) (mg/L N), nitrite (mg/L N),

nitrate (mg/L N), ortho-phosphate (mg/L P), TP

(mg/L P) and chlorophyll-a (Chl-a) (mg/m

Real-time physico-chemical data (pH, DO, conductivity

and temperature) were captured using the EXO

Sonde, which was deployed before sampling

(traditional boat and drone water sampling) and

recorded variables every second. On return to the lake

shore, data were downloaded to a laptop. The project

team used data produced from the nal 90 seconds of

the period in which the EXO Sonde was deployed in

the lake water to ensure that the probe had sufcient

time to adjust from recording data while in ight.

5.2 Comparison of Water Chemistry

Variables Collected Using

Traditional Boat and Drone

Water Sampling

Both sampling methodologies (traditional boat and

drone water sampling) collected 2 L of water on

100% of sampling occasions in addition to real-time

physico-chemical data for pH, DO, conductivity and

temperature.

For each water chemistry variable, each location

was included only if all three paired water samples

for each method exceeded the limits of detection.

Paired sample t-tests, for data meeting the

assumptions of parametric tests, indicated that

there were no signicant differences for alkalinity

[t = −0.416, degrees of freedom (df) = 9, p = 0.69, mean

difference = −1.27), hardness (t = 0.85, df = 5, p = 0.43,

mean difference = 1.67], true colour (t = −0.872, df = 11,

p = 0.41, mean difference = −0.78), silica (t = 0.89,

df = 11, p = 0.39, mean difference = 0.06), TON

(t = 0.775, df = 3, p = 0.5, mean difference = 0.002), TP

(t = 1.19, df = 4, p = 0.3, mean difference = 0.0005), Chl-a

(t = −1.99, df = 7, p = 0.09, mean difference = −0.25)

or conductivity (t = 1.89, df = 11, p = 0.09, mean

difference = 12.2) between traditional boat and drone

water sampling methodologies (Figure 5.1). The

Wilcoxon-signed rank test was used for data that were

non-normally distributed and/or had heterogeneous

variability. This indicated that there were no signicant

differences in the median concentrations of chloride

(Z = 0.614, df = 9, p = 0.54, mean difference = 0.04), DO

(Z = −0.63, df = 11, p = 0.53, mean difference = 0.056)

or temperature (Z = −0.94, df = 11, p = 0.35, mean

difference = −0.017) between traditional boat and

drone water sampling methodologies (Figure 5.1).

The only variable to show a signicant difference

between traditional boat and drone water sampling

methodologies was pH (t = −2.46, df = 11, p = 0.031,

mean difference = −0.048), although this is not deemed

hydrochemically or biologically important.

H. Lally et al. (2017-W-MS-28)

5.3 Comparison of Variability and

Precision in Water Chemistry

Variables Collected Using

Traditional Boat and Drone

Water Sampling Methodologies

A statistical assessment of variability, by calculating

the coefcient of variation for each of the three

repeated measurements at each sampling location

for each variable and each sampling methodology,

indicated that there was no signicant difference

(Z = −0.197, p = 0.85, average boat = 3.29%, average

drone = 3.76%) in overall variability between data

collected by drone and by boat. Moreover, the

precision of the data for each variable was also

statistically analysed separately for both water

sampling methodologies and only one signicant

difference, for hardness (Z = 2.87, p = 0.043, average

boat = 7.3%, average drone = 3.9%) (Table 5.2), was

found. Levels of hardness were found to be higher in

water samples taken using the traditional boat method

but are deemed within the range of detection.

Overall, natural variability and high precision in water

chemistry variable data were evident from water

samples taken using both the traditional boat and

drone water sampling methodologies. Thus, water

samples taken using the drone consistently matched

those of samples taken using traditional methods.

Table 5.1. Key characteristics of lakes sampled during eld trials

Lake

No. of

sampling

locations

Co-ordinates

for sampling

locations

Surface area

(ha)

Included

in the EPA

WFD Lake

Monitoring

Programme

WFD

alkalinity

status

WFD

typology

class

WFD status

Lough Fee 1 53.59122

−9.8381

174

– – –

Lough Inagh 1 53.5162

−9.73816

310

– – –

Lough Conn 2 53.9898

−9.25791

53.09365

−9.29682

4704



High 12 Moderate

Ballyquirke

Lough

2 53.32469

−9.15257

53.32603

−9.1543

73.6



Moderate 6 Bad

Lough Derg 3 52.90733

−8.50461

52.92032

−8.45241

52.91859

−8.45476

13,000



High 12 Moderate

Lough Mask 3 53.56779

−9.41073

53.56526

−9.41522

53.64387

−9.36527

8218



High 12 Good

Data taken from Inland Fisheries Ireland National Research Survey Programme Fish Stock Assessments 2015 and 2016

(Kelly et al., 2016, 2017a,b; McLoone et al., 2017).

Data taken from the EPA Maps portal (EPA, 2019).

Assessing the Potential of Drones to Take Water Samples and Physico-chemical Data from Open Lakes

Figure 5.1. Statistical comparison of water chemistry variable data collected using traditional boat

(x-axis) and drone (y-axis) water sampling. Legend: Lough Inagh, Lough Fee, Lough Conn 1,

Lough Conn 2, Ballyquirke Lough 1, Ballyquirke Lough 2, Lough Mask 1, Lough Mask 2,

Lough Mask 3, Lough Derg 1, Lough Derg 2 and Lough Derg 3.

H. Lally et al. (2017-W-MS-28)

5.4 Limitations Encountered during

Field Trials

Following completion of the eld trials, several

limitations regarding the use of the DJI M600 Pro

drone and attached water sampling prototype

payload were noted and require further consideration

(Table 5.3). A key concern is the operational capacity

of the DJI M600 Pro drone to carry an 8-kg payload.

This is beyond the specications indicated by DJI for

the drone platform (see Chapter 3, Table 3.1). Options

open to the team are to either reduce the weight of the

prototype payload or employ a larger drone, although

this would have knock-on cost implications.

Table 5.2. Statistical assessment of precision, using coefcient of variation, between water chemistry

variable data at each sampling location between traditional boat and drone water sampling

methodologies

Variable Test statistic No. of pairs p-value

Average CV (%)

Boat Drone

True colour

–1.54 12 0.15 2.9 4.3

Hardness 2.87 6 0.043 7.3 3.9

Silica

–0.89 12 0.4 7.6 8.7

TON

–0.19 4 0.86 1.3 1.6

Chloride –0.15 10 0.89 0.9 0.9

Alkalinity

–0.77 10 0.44 6.6 9.4

Chl-a –1.25 8 0.25 6.4 8.9

–0.41 5 0.69 6.9 5.9

–1.81 12 0.07 1.1 0.6

Temperature

–1.73 12 0.08 0.6 0.4

Conductivity

0.08 12 0.94 0.6 0.6

DO 0.2 12 0.84 0.5 0.5

Square root transformed.

Wilcoxon-signed rank test.

CV, coefcient of variation.

Table 5.3. List of limitations encountered during the current study when using the drone and water

sampling payload to collect water samples from Irish lakes

Limitations DJI M600 Pro drone with attached water sampling prototype payload

Weather Limited by wind speeds greater than 8 m/s

Limited by moderate to high rainfall levels

Limited number of suitable ying days in Ireland

Drone platform Operational carrying capacity of DJI M600 Pro drone is exceeded

Payload Payload

weight should be less than 6 kg

Payload weight increases operational risk of the drone (ability to lift payload)

Payload can swing below the drone in moderate to high winds making steady ight difcult

Payload weight affects ight times and this, in turn, can limit the distance between consecutive sampling

locations

Payload weight limits battery endurance where one set of six batteries is utilised per ight

Personnel Must have experienced drone pilots licensed with the IAA

Including weight of water sample.

6 Cost–BenetAnalysis

A cost–benet analysis of the project focused on time,

costs, resources, health and safety, and biosecurity

risks.

6.1 EfciencyofTraditionalBoat

versus Drone Water Sampling

Methods

At Ballyquirke Lough, on 29 November 2019, the

timings of all activities related to traditional boat and

drone water sampling methods were recorded (Table

6.1). All timings were calculated as person-minutes,

e.g. it took two people a total of 26 minutes 49 seconds

to wash the boat after sampling (for biosecurity

reasons), which equates to 53 minutes 38 seconds

in person-minutes. Using traditional boat water

sampling, it took 99 person-minutes to take three

replicate samples, while it took 45 person-minutes to

complete the same task using drone water sampling.

The time taken to clean and disinfect the boat was

54 person-minutes, while for the prototype payload

cleaning and disinfecting took 29 person-minutes.

Overall, where two persons are involved in undertaking

traditional boat and drone water sampling, the drone

is 2.3 times more efcient at capturing water samples

(Figure 6.1a); with three persons (e.g. two on board

the boat and one shore support person) involved

in undertaking traditional boat water sampling, a

two-person drone team is 3 times more efcient

(Figure 6.1b). In the future, it is envisaged that drone

water sampling could be conducted by only one

person, making drone water sampling methods 3.4

times more efcient than a two-person team conducting

traditional boat water sampling (Figure 6.1c).

6.2 Costs of Traditional Boat versus

Drone Water Sampling Methods

The capital costs associated with the boat water

sampling implemented by the project team were

Table 6.1. Timings of activities related to traditional boat versus drone water sampling at Ballyquirke

Lough

Traditional boat water sampling

Estimated

time Drone water sampling

Estimated

time

Activities related to sampling

Boat set-up 00:20:41 Drone set-up 00:00:54

Initial set-up of PC and YSI Sonde and probes 00:06:17 Initial set-up of PC and YSI Sonde and probes 00:06:17

Bottle labelling 00:03:13 Bottle labelling 00:03:13

Table set-up

00:01:10

Boat sampling (from leaving shore to return and

completion of data download)

00:10:00 Drone sampling (from leaving shore to return and

completed data download)

00:11:04

Boat disassembly 00:09:10

Total sampling time 00:49:21 Total sampling time 00:22:38

Person-minutes per sampling location

a,b

01:38:42 Person-minutes per sampling location

a,b

00:45:16

Biosecurity measures

Boat cleaning including disinfecting PPE, mooring,

ropes, oars, etc.

00:26:49 Drone and payload cleaning including disinfecting

payload

00:11:03

Person-minutes to disinfect equipment

00:53:38 Person-minutes to disinfect equipment

00:22:06

Total person-minutes per boat sampling location

a,b

02:32:20 Total person-minutes per drone sampling location

a,b

01:07:22

Sampling location refers to a lake site where three replicate samples were collected using either traditional boat or drone

water sampling methods.

Two persons were involved in sampling.

Foldable table utilised by the drone team for changing batteries and YSI Sonde and probes on the prototype payload.

PPE, personal protective equipment.

H. Lally et al. (2017-W-MS-28)

compared with those of using the drone and prototype

payload designed and developed during the project.

Capital costs associated with boat water sampling

employed during eld trials were those of an inatable

Zodiac Classic Mark II (including the costs of oars

and pump) and YSI Sonde and probes, which totalled

€21,286.06 (Table 6.2). It is important to note that

our capital cost comparison was made using the

Zodiac Classic Mark II as a demonstrative comparison

and therefore institutes and companies that utilise

different types of boats should bear this in mind when

evaluating the relative capital costs of a drone system

versus a boat for water sampling. Similarly, the number

of samples that can be collected in the lifetime of a

boat varies considerably depending on the make,

model, frequency of use and maintenance history. The

capital investment costs associated with the drone

water sampling employed during eld trials were those

of the DJI M600 Pro, batteries, undercarriage and

communication accessories, payload build and YSI

Sonde and probes, totalling €26,052.79.

The project team estimate that the DJI M600 Pro is

capable of conducting 500 water sampling missions

before renewal and so estimates for both the Zodiac

inatable boat and the DJI M600 Pro were based

on 500 water sampling missions. Therefore, boat

water sampling costs €42.57 per sample compared

with €52.11 per sample for the drone water sampling

method developed by the project team, making boat

water sampling 1.2 times cheaper per sample based

on capital investment costs.

However, traditional boat water sampling would

typically employ the use of handheld probes such as

a YSI Handheld Proplus with three probes and data-

logging capabilities. This reduces the total capital costs

associated with boat sampling further, to €9544.25, or

€19.09 per sample, making traditional boat sampling

2.7 times cheaper per sample based on capital

investment costs.

The capital costs associated with the DJI M600 Pro

and prototype payload build are relatively high but

Figure 6.1. Efciency of traditional boat versus drone water sampling under different resource scenarios.

(a) (b)

(c)

Assessing the Potential of Drones to Take Water Samples and Physico-chemical Data from Open Lakes

expected for such new and innovative technology.

The YSI Sonde and probe system is particularly costly

and, while allowing the transmission of real-time data

back to the ground station, may not be necessary

for every sampling mission. Therefore, the project

team proposes some alternatives at the current time.

First, the YSI Sonde and probes could be removed.

Second, pH, DO, conductivity and temperature could

be measured in situ using a YSI handheld device

(YSI Handheld Proplus) when water samples are

returned to the shore following drone sampling. This

would result in a cost of €28.62 per sample. However,

this is not ideal for measuring parameters such as

temperature and DO concentrations, although it is

unlikely that temperature and DO concentrations would

alter signicantly or in an ecologically meaningful

way in the time between collection and shore

measurement. Alternatively, the team considered not

conducting in situ physico-chemical measurements at

all, which would reduce costs to €22.53 per sample

(Table 6.2); however, this is undesirable for sampling

conducted under many monitoring programmes such

as the WFD monitoring programme. Therefore, the

project team suggests replacing the YSI Sonde with

a cheaper model to signicantly reduce the capital

outlay of the prototype payload. However, such an

alternative model would have to undergo experimental

trials to ensure that deployment via a drone would not

signicantly affect the data collected. Traditional boat

water sampling remains 1.2 to 1.5 times cheaper than

drone water sampling at the current time, although the

project team expect the capital costs associated with

the prototype payload to reduce if commercialised,

as payload manufacturing costs are streamlined,

reducing costs to be in line with those of traditional

boat sampling.

Overall, the capital costs currently associated with

traditional boat sampling are lower than those of drone

water sampling. However, drone water sampling is far

more efcient than the current traditional boat water

sampling method. Therefore, there is a balance to be

struck between the capital costs involved in setting

up a new water sampling methodology and the longer

term benets of drone water sampling in reducing the

much more signicant labour costs, increasing the

efciency of sampling, reducing the use of resources,

and decreasing health and safety and biosecurity risks.

6.3 Advantages and Disadvantages

of Traditional Boat and Drone

Water Sampling Methodologies

Following the completion of eld trials and the cost–

benet analysis, the project team compiled a list of

advantages and disadvantages requiring consideration

when choosing either the traditional boat or drone

water sampling methods (Table 6.3).

Key advantages of the drone water sampling method

are (1) the efciency of water sampling; (2) the ability

to access water bodies in remote and inaccessible

locations where boat access is unavailable; (3) the

ease of transporting the equipment; (4) lower

biosecurity risks and cleaning requirements;

Table 6.2. Capital costs associated with boat and drone water sampling during this research project

Estimated

capital costs

of boat water

sampling during

project eld

trials (€)

Traditional boat

sampling (€)

Estimated

capital costs

of drone water

sampling during

project eld

trials (€)

Drone water

sampling with

in situ handheld

physico-

chemical data

collection (€)

Drone water

sampling

with no in

situ physico-

chemical data

collection (€)

Inatable Zodiac (including

oars, pump and engine)

6500 6500 – – –

DJI M600 Pro platform – – 6811.95 6811.95 6811.95

Batteries – – 1102.08 1102.08 1102.08

Undercarriage and

communication accessories

– – 2539.89 2539.89 2539.89

Prototype payload build – – 812.82 812.82 812.82

YSI Sonde and probes 14,786.06 – 14,786.06 – –

YSI Handheld Proplus – 3044.25 – 3044.25 –

Cost per 500 samples 21,286.06 9544.25 26,052.79 14,310.98 11,266.73

Cost per sample 42.57 19.09 52.11 28.62 22.53

H. Lally et al. (2017-W-MS-28)

(5) the onboard camera can also survey for

hydromorphological and biological water features; and

(6) the increase in the number of unmonitored lakes

that can be surveyed in a session.

The advantages of the boat water sampling method

are that (1) it is more cost-effective in relation to capital

outlay; (2) it has the ability to sample large areas of

a lake; and (3) there are no weight restrictions on

equipment, although additional costs can be incurred if

boats are rented.

6.4 Views Gathered from End-of-

project Workshop Focus Groups

The project team gathered the views of participants

attending the end-of-project workshop, held at

GMIT on 25 March 2020, on several aspects of the

research ndings including drone water sampling, cost

effectiveness, drone legislation, and health and safety

and biosecurity risks (Table 6.4).

Table 6.3. Advantages and disadvantages of traditional boat and drone water sampling methodologies

Traditional boat water sampling Drone water sampling

Advantages

Efcient sampling method



Access to water bodies Limited to sites with slipways Safer access to remote and inaccessible

sites

Transportation Large boats can be cumbersome to

transport

Ease of transport between sites and set-up

Relatively low biosecurity risks



Onboard camera capable of capturing aerial

footage of additional physical and biological

features of the water body



Potential to increase the number of

monitored lakes sampled in a session



Cost-effective



Costly during project set-up

Ability to sample large areas of the lake Shallow waters and rocks can impede

access to sampling locations by boats

Require multiple ground stations or SOPs to

conduct ights BVLOS on large lakes

No weight restrictions on carrying capacity



Dependent on specication of drone to be

deployed

100% successful water capture

 

Ability to sample up to 300 m offshore

 

Reliable and accurate water chemistry

measurements

 

No cross-contamination between samples

 

Drone platforms are affordable



Disadvantages

Potential long-term impact on job security



Additional boat hire costs



Weather Limited by wind speeds greater than

10 m/s

Limited by wind speeds greater than 8 m/s

and moderate or heavy rainfall which could

delay or cancel UAV operations

Cost of replacement parts High (including the costs of the boat,

boat trailer, boat repair)

High (including the costs of batteries,

electronic equipment)

Real-time data capture High cost of water sampling probes

with real-time data transfer capabilities

High cost of water sampling probes with

real-time data transfer capabilities

Requires adequate space and secure

storage

û û

Workplace health and safety concerns

û û

High insurance costs

û û

SOP, special operating permission.

[AQ1]

Assessing the Potential of Drones to Take Water Samples and Physico-chemical Data from Open Lakes

6.4.1 Drone water sampling

Overall, the focus groups indicated that they would

consider the use of drone water sampling in the future

but overwhelmingly felt that drone water sampling,

in its current format, would not be able to completely

replace traditional boat sampling methods (Figures 6.2

and 6.3). Rather, a tandem sampling approach was

favoured that would complement sampling at remote

and inaccessible locations, and where water chemistry

data only were required.

6.4.2 Cost-effectiveness

Participants of the focus groups felt that the capital

costs associated with drone water sampling were

not good value for money. In addition, if drone water

sampling was requested via a tender process the costs

would be far beyond those of traditional boat sampling

for many suppliers of such services. Thus, participants

deemed the prototype drone water sampling method

developed by the project team to be not competitive or

not commercially or economically viable in its current

format. They did, however, suggest a shared service

as a means by which several agencies could come

together and share the capital costs and benets of

drone water sampling, which would reduce the cost of

water sampling for all.

In-house training would be deemed suitable if routine

drone water sampling and shared services were to be

undertaken by local authority agencies and permanent

staff within the EPA but not for small consultancies or

water testing laboratories.

6.4.3 Drone legislation

Overall, knowledge of drone legislation was limited

and restricted to knowledge of the maximum height

(120 m), maximum distance (300 m) and weight of

Table 6.4. List of questions posed to participants during the focus group session at the end-of-project

workshop held at GMIT on 25 March 2020

Topic Question

Drone water sampling

Q1 Would you consider drone water sampling as a sampling methodology in the

future?

If so, for what purposes would you apply the system?

If not, why?

Q2 What do you consider to be the major benets of applying such methodologies in

your workplace?

Q3 What would be the major considerations/constraints with applying such a

methodology in your workplace?

Q4 Would the methodology replace your current water sampling protocols or would

you operate tandem methodologies?

Cost-effectiveness

Q1 How much would you be willing to invest in this new methodology if it were to be

implemented in your workplace?

Q2 Would you consider in-house training of personnel to conduct ight operations or

would you tender a contract to external drone operators?

Drone legislation

Q1 What is your knowledge of drone legislation in Ireland?

Q2 Do you know how drone legal requirements are applied within the workplace when

conducting aquatic/terrestrial surveys using drones as part of your work/research?

Health and safety

Q1 In your workplace, would the drone water sampling method offer a safer working

environment?

Q2 Would this reduce the associated insurance and liability costs for your employer?

Biosecurity risk

Q1 How would you rate the biosecurity risk of the drone water sampling prototype

device?

Low, medium or high

Why?

Q2 How does this compare to the biosecurity risk associated with the boat?

H. Lally et al. (2017-W-MS-28)

drones requiring registration with the IAA and distance

required from 12 people or more (120 m).

In contrast, participants had a clearer understanding of

drone rules and regulations when applied to their work/

research, with many groups highlighting additional

requirements for land owner/local authority permission;

no y zones around national parks, Special Protection

Areas, Special Areas of Conservation, airports,

aerodromes and Ofce of Public Works lands; that

drones must operate within the visual line of sight and

be added to company public limited liability insurance;

and that pilots should have undertaken and completed

ground school training with an approved provider.

6.4.4 Health and safety

All focus groups agreed that drone water sampling

was safer for staff but that this alone would not reduce

associated insurance or liabilities for their employers.

In situations where employers operated both boat

and drone water sampling in tandem, the combined

insurance could increase liabilities.

6.4.5 Biosecurity risk

All focus groups deemed drone water sampling to be

a lower biosecurity risk than boats and to be easier

and to require less people power to clean and disinfect

effectively.

Reduced health and

safety risks

Access to remote and

inaccessible water

bodies

Increased sampling of

unmonitored WFD

lakes

Efficient sampling

method

Capture of more in

situ, real-time data

Obtain accurate and

reproducible data

Adaptability of drone

technology to

conduct water

sampling

Provide a more

inclusive picture of

lake water quality in

Ireland

Figure 6.2. Benets of drone water sampling highlighted by participants during the focus group session

at the end-of-project workshop held at GMIT on 25 March 2020.

Loss of assets Weather dependent

Costs dependent on

scale of project

Availability of water

sampling payloads

Requirement for

pilot training

Social/ethical and

regulatory

restrictions

Maximum distance

drone can fly and

take samples

Limited battery

endurance

Accreditation of in

situ data

Limited sample size

(2 L)

Figure 6.3. Limitations of drone water sampling highlighted by participants during the focus group

session at the end-of-project workshop held at GMIT on 25 March 2020.

7 Key Findings

This research has made signicant contributions

to (1) the advancement of drone water sampling

technology, in particular the design and build of a

prototype payload to consistently conduct water

sampling; (2) the design of robust comparative

experimental eld trials that clearly demonstrate

that the prototype payload designed collects both

hydrochemical data and results from laboratory-tested

water samples that are the same as those collected

via traditional boat sampling; (3) conducting the rst

and informative cost–benet analysis of the use of

drones to collect lake water hydrochemical data and

samples; and (4) publishing the rst review of drone

water sampling techniques used worldwide.

7.1 Achievements of this Research

Key achievements of this research:

● This was the rst research team, worldwide, to

publish a critical review of drone water sampling

techniques.

● The drone and attached prototype payload, as

demonstrated during eld trials, can be used to

successfully collect water samples from open

lakes.

● The drone and attached prototype payload can

successfully collect water samples from 100 m

offshore.

● The water sampling time achieved using the drone

and attached prototype payload is 4 minutes.

● The prototype payload developed by the project

team is the rst in Europe to collect 2 L of water

using a drone.

● The water sampling rates are 100%.

● Real-time physico-chemical data capture is

possible and allows users to review data before

leaving a site.

● The comparison of a wide range of water

chemistry variables showed no signicant

differences between traditional boat and drone

water sampling methods.

● Precision was not signicantly affected by the

sampling methodology employed.

● A comparative analysis of water chemistry

variables satises the requirements of lake water

sampling and monitoring in Ireland under the

EU WFD.

● The capital costs associated with water sampling

methods were considered through a cost–benet

analysis as part of the project.

● The capital investment costs of traditional boat

sampling were found to be 1.2 to 1.5 times lower

than those of drone water sampling but, more

importantly, drone water sampling was found to

be 2.3 to 3.4 times faster than traditional boat

sampling methods, depending on resource

allocation.

● Drone water sampling reduces the health and

safety and biosecurity risks associated with open

lake sampling.

● Drone water sampling offers a unique solution

to sampling in remote and inaccessible lakes

and unmonitored lakes under the WFD, and to

conrm the water quality of lakes categorised

using remote-sensing methods, increasing our

understanding of water quality in Irish lakes.

● The application of drone water sampling for

large-scale water sampling programmes has been

proven and can be adapted as needed to other

aquatic environments.

8 Recommendations

The project team makes the following

recommendations to further the advancement of drone

water sampling over the coming years:

● Consideration should be given to deploying

waterproof drones such as the Freey Alta 8 or

Alta X (see Chapter 3, Table 3.1). These drones

would allow ights to operate in less than optimal

weather conditions, including in wind speeds

greater than 8 m/s and moderate rainfall.

● Smaller, lighter and cheaper real-time water

chemistry probes should be integrated, to reduce

the weight of the payload and the associated

capital costs required during project set-up.

This would, in turn, avoid the need for larger,

more expensive drones and would allow users

to operate within evolving legislative and pilot

training requirements.

● The design and build of the prototype payload,

especially its size (height and length) and weight,

should be rened to allow a greater number of

off-the-shelf drones to be considered for use.

The use of additive fabrication techniques, which

would allow the printing of new or replacement

parts when needed or damaged, should also be

considered.

● Field trials should continue and should include a

wider variety of aquatic environments including

estuaries at high and low tides, marinas

and streams and rivers at various ows, to

demonstrate application across various aquatic

environments.

References

Castendyk, D., Hill, B., Filiatreault, P., Straight, B.,

Alangari, A., Cote, P. and Leishman, W., 2018.

Experiences with autonomous sampling of pit lakes in

North America using drone aircraft and drone boats. In

Wolkersdorfer, C., Sartz, L., Weber, A., Burgess, J. and

Tremblay, G. (eds), Mine Water – Risk to Opportunity.

Tshwane University of Technology, Pretoria, South

Africa, pp. 1036–1042.

Castendyk, D., Straight, B.J., Voorhis, J.C., Somogyi,

M.K., Jepson, W.E. and Kucera, B.L., 2019. Using

aerial drones to select sample depths in pit lakes. In

Fourie, A.B. and Tibbett, M. (eds), Mine Closures.

Case Studies and Remote Survey. Australian Centre

for Geomechanics, Perth, Australia.

Castendyk, D., Voorhis, J. and Kucera, B., 2020. A

validated method for pit lake water sampling using

aerial drones and sampling devices. Mine Water and

the Environment 39: 440–454. http://doi.org/10.1007/

s10230-020-00673-y

Chabot, D., 2018. Trends in drone research and

applications as the Journal of Unmanned Vehicle

Systems turns ve. Journal of Unmanned Vehicle

Systems 6(1): vi–xv. https://doi.org/10.1139/

juvs-2018-0005

Chapman, A., 2014. It’s okay to call them drones. Journal

of Unmanned Vehicle Systems 2(2): iii–v. http://doi.

org/10.1139/juvs-2014-0009

Detweiler, C., Ore, J.-P., Anthony, D., Elbaum, S., Burgin,

A. and Lorenz, A., 2015. Bringing unmanned aerial

systems closer to the environment. Environmental

Practice 17: 188–200. https://doi.org/10.1017/

S1466046615000174

EASA (European Aviation Safety Agency), 2015. Riga

Declaration on Remotely Piloted Aircraft (Drones).

EASA, Riga.

EASA (European Aviation Safety Agency), 2018. Opinion

No 1/2018 – Introduction of a regulatory framework

for the operation of unmanned aircraft systems in the

“open” and “specic” categories. EASA, Cologne,

Germany.

EPA (Environmental Protection Agency), 2006. Water

Framework Directive Monitoring Programme. EPA,

Johnstown Castle, Ireland.

EPA (Environmental Protection Agency), 2007. Water

Framework Directive: Proposed Quality Standards for

Surface Water Classication – A Discussion Document

for Public Consultation. EPA, Johnstown Castle,

Ireland.

EPA (Environmental Protection Agency), 2011. Water

Framework Status Update Based on Monitoring

Results 2007–2009. EPA, Johnstown Castle, Ireland.

EPA (Environmental Protection Agency), 2019. EPA

Maps. Available online: https://gis.epa.ie/EPAMaps/

EU (European Union), 2000. Directive 2000/60/EC of the

European Parliament and of the Council of 23 October

2000 establishing a framework for Community action

in the eld of water policy. OJ L 327, 22.12.2000,

pp. 1–73.

Fráter, T., Juzsakova, T., Lauer, J., Dióssy, L. and Rédey,

A., 2015. Unmanned aerial vehicles in environmental

monitoring – an efcient way for remote sensing.

Journal of Environmental Science and Engineering A 4:

85–91. http://doi.org/10.17265/2162-5298/2015.02.004

Free, G., Little, R., Tierney, D., Donnelly, K. and Caroni,

R., 2006. A Reference-based Typology and Ecological

Assessment System for Irish Lakes: Preliminary

Investigations – Final Report. Environmental

Protection Agency, Johnstown Castle, Ireland.

Free, G., Little, R., Tierney, D., Donnelly, K. and Caroni,

R., 2007. Water Framework Directive – A Reference-

based Typology and Ecological Assessment

System for Irish Lakes: Preliminary Investigations.

Environmental Protection Agency, Johnstown Castle,

Ireland.

Free, G., Bresciani, M., Trodd, W., Tierney, D., O’Boyle,

S., Plant, C. and Deakin, J., 2020. Estimation of lake

ecological quality from Sentinel-2 remote sensing

imagery. Hydrobiologia 847: 1423–1438. https://doi.

org/10.1007/s10750-020-04197-y

Hering, D., Borja, A., Carstensen, J., Carvalho, L., Elliott,

M., Feld, C.K., Heiskanen, A.-S., Johnson, R.K., Moe,

J., Pont, D., Solheim, A. L. and van de Bund, W.,

2010. The European Water Framework Directive at the

age of 10: a critical review of the achievements with

recommendations for the future. Science of the Total

Environment 408: 4007–4019.

ICAO (International Civil Aviation Organization), 2011.

Unmanned Aircraft Systems (UAS). ICAO, Montreal,

QC, Canada.

JARUS (Joint Authorities for Rulemaking of Unmanned

Systems), 2013. Certication Specication for Light

Unmanned Rotorcraft Systems (CS-LURS). Available

online: http://jarus-rpas.org/sites/jarus-rpas.org/

les/storage/Library-Documents/jar_01_doc_jarus_

certication_specication_for_lurs_-_30_oct_2013.pdf

H. Lally et al. (2017-W-MS-28)

Kelly, F.L., Connor, L., Delanty, K., McLoone P., Coyne, J.,

Morrissey, E., Corcoran, W., Cierpial, D., Matson, R.,

Gordon, P., O’Briain, R., Rocks, K., Walsh, L., O’Reilly,

S., O’Callaghan, R., Cooney, R. and Timbs, D., 2016.

Fish Stock Survey of Lough Mask, June 2015. National

Research Survey Programme, Inland Fisheries

Ireland, Dublin.

Kelly, F.L., Connor, L., Morrissey, E., McLoone, P., Kelly,

K., Delanty, K., Coyne, J., Corcoran, W., Cierpial,

D., Matson, R., Gordon, P., O’Briain, R., Rocks, K.,

O’Reilly, S., Puttharee, D., McWeeney, D., Robson,

S. and Buckley, S., 2017a. Fish Stock Survey of

Lough Derg, June 2016. National Research Survey

Programme, Inland Fisheries Ireland, Dublin.

Kelly, F.L., Connor, L., Delanty, K., McLoone, P., Coyne,

J., Morrissey, E., Corcoran, W., Cierpial, D., Matson,

R., Gordon, P., O’Briain, R., Rocks, K., O’Reilly, S.,

Kelly, K., Puttharee, D., McWeeney, D., Robson, S.

and Buckley, S., 2017b. Fish Stock Survey of Lough

Conn, August 2016. National Research Survey

Programme, Inland Fisheries Ireland, Dublin.

Koparan, C. and Koc, A.B., 2016. Unmanned aerial

vehicle (UAV) assisted water sampling. Paper

presented at the 2016 American Society of

Agricultural and Biological Engineers (ASABE) Annual

International Meeting, Orlando, Florida, 17–20 July.

http://doi.org/10.13031/AIM.20162461157

Koparan, C., Koc, A.B., Privette, C.V. Sawyer, C.B. and

Sharp, J.L., 2018a. Evaluation of a UAV-assisted

autonomous water sampling. Water 10(5): 655–681.

http://doi.org/10.3390/w10050655

Koparan, C., Koc, A.B., Privette, C.V. and Sawyer, C.B.,

2018b. In situ water quality measurements using an

unmanned aerial vehicle (UAV) system. Water 10(3):

264–278. http://doi.org/10.3390/w10030264

Koparan, C., Koc, A.B., Privette, C.V. and Sawyer,

C.B., 2019. Autonomous in situ measurements of

noncontaminant water quality indicators and sample

collection with a UAV. Water 11(3): 604–619. http://doi.

org/10.3390/w11030604

Koparan, C., Koc, A.B., Privette, C.V. and Sawyer, C.B.,

2020. Adaptive water sampling device for aerial robots.

Drones 4: 5–21. http://doi.org/10.3390/drones4010005

McLoone, P., Connor, L., Coyne, J., Morrissey, E.,

Corcoran, W., Cierpial, D., Delanty, K., Matson, R.,

Gordon, P., O’Briain, R., Rocks, K., O’Reilly, S.,

Puttharee, D., McWeeney, D., Robson, S., Buckley, S.

and Kelly, F.L., 2017. Fish Stock Survey of Ballyquirke

Lough, September 2016. National Research Survey

Programme, Coarse Fish and Pike, Inland Fisheries

Ireland, Dublin.

O’Boyle, S., Trodd, W., Bradley, C., Tierney, D., Wilkes,

R., Ní Longphuirt, S., Smith, J., Stephens, A., Barry,

J., Maher, P., McGinn, R., Mockler, E., Deakin, J.,

Craig, M. and Gurrie, M., 2019. Water Quality in

Ireland 2013–2018. Environmental Protection Agency,

Johnstown Castle, Ireland.

Ore, J.-P., Elbaum, S., Burgin, A., Zhao, B. and Detweiler,

C., 2013. Autonomous aerial water sampling. Field

and Service Robots: Results of the 9

International

Conference, Brisbane, 9–11 December, Australia,

pp. 137–151.

Ore, J.-P., Elbaum, S., Burgin, A. and Detweiler, C., 2015.

Autonomous aerial water sampling. Journal of Field

Robotics 32(8): 1095–1113. https://doi.org/10.1002/

rob.21591

Song, K., Brewer, A., Ahmadian, S., Shankar, A.,

Detweiler, C. and Burgin, A., 2017. Using unmanned

aerial vehicles to sample aquatic ecosystems.

Limnology and Oceanography: Methods 15: 1021–

1030. http://doi.org/10.1002/lom3.10222

Terada, A., Morita, Y., Hashimoto, T., Mori, T., Ohba, T.,

Yaguchi, M. and Kanda, W., 2018. Water sampling

using a drone at Yugama crater lake, Kusatsu-Shirane

volcano, Japan. Earth, Planets and Space 70(64): 1–9.

https://doi.org/10.1186/s40623-018-0835-3

Tierney, D., Free, G., Kennedy, B., Little, R., Plant, C.,

Trodd, W. and Wynne, C., 2015. Lakes. In Byrne, C.

and Fanning, A. (eds), Water Quality in Ireland 2010–

2012. Environmental Protection Agency, Johnstown

Castle, Ireland, pp. 90–108.

US EPA (United States Environmental Protection

Agency), 1994. Laboratory data validation functional

guidelines for evaluating inorganics analyses.

Publication 9240. Ofce of Solid Waste and

Emergency Response, US EPA, Washington, DC,

pp.1–26.

Vergouw, B., Nagel, H., Bondt, G. and Custers, B., 2016.

Drone technology: types, payloads, applications,

frequency spectrum issues and future developments.

In Custers, B. (ed.), The Future of Drone Use.

Information Technology and Law Series, vol. 27,

T.M.C. Asser Press, The Hague, the Netherlands,

pp. 21–45. http://doi.org/10.1007/978-94-6265-132-6_2

Wich, S.A. and Koh, L.P., 2018. Conservation Drones:

Mapping and Monitoring Biodiversity. Oxford University

Press, New York, NY.

Wynne, C. and Donohoe, I., 2016. Predicting Ecological

Status in Unmonitored Lakes Using Catchment

Land Use and Hydromorphological Characteristics.

Environmental Protection Agency, Johnstown Castle,

Ireland.

Abbreviations

BVLOS Beyond visual line of sight

Chl-a Chlorophyll-a

CRDL Contract-required detection limit

CTD Conductivity, temperature and depth

df Degrees of freedom

DO Dissolved oxygen

EPA Environmental Protection Agency

EU European Union

GMIT Galway-Mayo Institute of Technology

GPS Global positioning system

HDPE High-density polyethylene

IAA Irish Aviation Authority

MFRC Marine and Freshwater Research Centre

RPD Relative percentage difference

TON Total oxidised nitrogen

TP Total phosphorus

UAV Unmanned aerial vehicle

WFD Water Framework Directive

AN GHNÍOMHAIREACHT UM CHAOMHNÚ COMHSHAOIL

Tá an Ghníomhaireacht um Chaomhnú Comhshaoil (GCC) freagrach as an

gcomhshaol a chaomhnú agus a fheabhsú mar shócmhainn luachmhar do

mhuintir na hÉireann. Táimid tiomanta do dhaoine agus don chomhshaol a

chosaint ó éifeachtaí díobhálacha na radaíochta agus an truaillithe.

Is féidir obair na Gníomhaireachta a

roinnt ina trí phríomhréimse:

Rialú: Déanaimid córais éifeachtacha rialaithe agus comhlíonta

comhshaoil a chur i bhfeidhm chun torthaí maithe comhshaoil a

sholáthar agus chun díriú orthu siúd nach gcloíonn leis na córais sin.

Eolas: Soláthraímid sonraí, faisnéis agus measúnú comhshaoil atá

ar ardchaighdeán, spriocdhírithe agus tráthúil chun bonn eolais a

chur faoin gcinnteoireacht ar gach leibhéal.

Tacaíocht: Bímid ag saothrú i gcomhar le grúpaí eile chun tacú

le comhshaol atá glan, táirgiúil agus cosanta go maith, agus le

hiompar a chuirdh le comhshaol inbhuanaithe.

Ár bhFreagrachtaí

Ceadúnú

Déanaimid na gníomhaíochtaí seo a leanas a rialú ionas nach

ndéanann siad dochar do shláinte an phobail ná don chomhshaol:

• saoráidí dramhaíola (m.sh. láithreáin líonta talún, loisceoirí,

stáisiúin aistrithe dramhaíola);

• gníomhaíochtaí tionsclaíocha ar scála mór (m.sh. déantúsaíocht

cógaisíochta, déantúsaíocht stroighne, stáisiúin chumhachta);

• an diantalmhaíocht (m.sh. muca, éanlaith);

• úsáid shrianta agus scaoileadh rialaithe Orgánach

Géinmhodhnaithe (OGM);

• foinsí radaíochta ianúcháin (m.sh. trealamh x-gha agus

radaiteiripe, foinsí tionsclaíocha);

• áiseanna móra stórála peitril;

• scardadh dramhuisce;

• gníomhaíochtaí dumpála ar farraige.

Forfheidhmiú Náisiúnta i leith Cúrsaí Comhshaoil

• Clár náisiúnta iniúchtaí agus cigireachtaí a dhéanamh gach

bliain ar shaoráidí a bhfuil ceadúnas ón nGníomhaireacht acu.

• Maoirseacht a dhéanamh ar fhreagrachtaí cosanta comhshaoil na

n-údarás áitiúil.

• Caighdeán an uisce óil, arna sholáthar ag soláthraithe uisce

phoiblí, a mhaoirsiú.

• Obair le húdaráis áitiúla agus le gníomhaireachtaí eile chun dul

i ngleic le coireanna comhshaoil trí chomhordú a dhéanamh ar

líonra forfheidhmiúcháin náisiúnta, trí dhíriú ar chiontóirí, agus

trí mhaoirsiú a dhéanamh ar leasúchán.

• Cur i bhfeidhm rialachán ar nós na Rialachán um

Dhramhthrealamh Leictreach agus Leictreonach (DTLL), um

Shrian ar Shubstaintí Guaiseacha agus na Rialachán um rialú ar

shubstaintí a ídíonn an ciseal ózóin.

• An dlí a chur orthu siúd a bhriseann dlí an chomhshaoil agus a

dhéanann dochar don chomhshaol.

Bainistíocht Uisce

• Monatóireacht agus tuairisciú a dhéanamh ar cháilíocht

aibhneacha, lochanna, uiscí idirchriosacha agus cósta na

hÉireann, agus screamhuiscí; leibhéil uisce agus sruthanna

aibhneacha a thomhas.

• Comhordú náisiúnta agus maoirsiú a dhéanamh ar an gCreat-

Treoir Uisce.

• Monatóireacht agus tuairisciú a dhéanamh ar Cháilíocht an

Uisce Snámha.

Monatóireacht, Anailís agus Tuairisciú ar

an gComhshaol

• Monatóireacht a dhéanamh ar cháilíocht an aeir agus Treoir an AE

maidir le hAer Glan don Eoraip (CAFÉ) a chur chun feidhme.

• Tuairisciú neamhspleách le cabhrú le cinnteoireacht an rialtais

náisiúnta agus na n-údarás áitiúil (m.sh. tuairisciú tréimhsiúil ar

staid Chomhshaol na hÉireann agus Tuarascálacha ar Tháscairí).

Rialú Astaíochtaí na nGás Ceaptha Teasa in Éirinn

• Fardail agus réamh-mheastacháin na hÉireann maidir le gáis

cheaptha teasa a ullmhú.

• An Treoir maidir le Trádáil Astaíochtaí a chur chun feidhme i gcomhair

breis agus 100 de na táirgeoirí dé-ocsaíde carbóin is mó in Éirinn.

Taighde agus Forbairt Comhshaoil

• Taighde comhshaoil a chistiú chun brúnna a shainaithint, bonn

eolais a chur faoi bheartais, agus réitigh a sholáthar i réimsí na

haeráide, an uisce agus na hinbhuanaitheachta.

Measúnacht Straitéiseach Timpeallachta

• Measúnacht a dhéanamh ar thionchar pleananna agus clár beartaithe

ar an gcomhshaol in Éirinn (m.sh. mórphleananna forbartha).

Cosaint Raideolaíoch

• Monatóireacht a dhéanamh ar leibhéil radaíochta, measúnacht a

dhéanamh ar nochtadh mhuintir na hÉireann don radaíocht ianúcháin.

• Cabhrú le pleananna náisiúnta a fhorbairt le haghaidh éigeandálaí

ag eascairt as taismí núicléacha.

• Monatóireacht a dhéanamh ar fhorbairtí thar lear a bhaineann le

saoráidí núicléacha agus leis an tsábháilteacht raideolaíochta.

• Sainseirbhísí cosanta ar an radaíocht a sholáthar, nó maoirsiú a

dhéanamh ar sholáthar na seirbhísí sin.

Treoir, Faisnéis Inrochtana agus Oideachas

• Comhairle agus treoir a chur ar fáil d’earnáil na tionsclaíochta

agus don phobal maidir le hábhair a bhaineann le caomhnú an

chomhshaoil agus leis an gcosaint raideolaíoch.

• Faisnéis thráthúil ar an gcomhshaol ar a bhfuil fáil éasca a

chur ar fáil chun rannpháirtíocht an phobail a spreagadh sa

chinnteoireacht i ndáil leis an gcomhshaol (m.sh. Timpeall an Tí,

léarscáileanna radóin).

• Comhairle a chur ar fáil don Rialtas maidir le hábhair a

bhaineann leis an tsábháilteacht raideolaíoch agus le cúrsaí

práinnfhreagartha.

• Plean Náisiúnta Bainistíochta Dramhaíola Guaisí a fhorbairt chun

dramhaíl ghuaiseach a chosc agus a bhainistiú.

Múscailt Feasachta agus Athrú Iompraíochta

• Feasacht chomhshaoil níos fearr a ghiniúint agus dul i bhfeidhm

ar athrú iompraíochta dearfach trí thacú le gnóthais, le pobail

agus le teaghlaigh a bheith níos éifeachtúla ar acmhainní.

• Tástáil le haghaidh radóin a chur chun cinn i dtithe agus in ionaid

oibre, agus gníomhartha leasúcháin a spreagadh nuair is gá.

Bainistíocht agus struchtúr na Gníomhaireachta um

Chaomhnú Comhshaoil

Tá an ghníomhaíocht á bainistiú ag Bord lánaimseartha, ar a bhfuil

Ard-Stiúrthóir agus cúigear Stiúrthóirí. Déantar an obair ar fud cúig

cinn d’Oigí:

• An Oig um Inmharthanacht Comhshaoil

• An Oig Forfheidhmithe i leith cúrsaí Comhshaoil

• An Oig um Fianaise is Measúnú

• Oig um Chosaint Radaíochta agus Monatóireachta Comhshaoil

• An Oig Cumarsáide agus Seirbhísí Corparáideacha

Tá Coiste Comhairleach ag an nGníomhaireacht le cabhrú léi. Tá

dáréag comhaltaí air agus tagann siad le chéile go rialta le plé a

dhéanamh ar ábhair imní agus le comhairle a chur ar an mBord.

www.epa.ie

Idenfying Pressures

Water sampling remains a key component in the monitoring and assessment of aquac environments. Sampling

requiring the use of a boat can lead to issues around accessibility, parcularly at remote lakes where there may be a

lack of a slipway. In addion, there are considerable cost implicaons, mainly related to the boat costs, the need for

dierent-sized boats for dierent lakes and the signicant me and resource requirements. Boat sampling can also

pose many health and safety issues, as well as biosecurity risks, which can be exacerbated in remote regions where

access is dicult. The applicaon of drones to collect in situ hydrochemical data and retrieve water samples from

freshwater environments provides the potenal to full some aspects of the biological and physicochemical sampling

required to meet large-scale water sampling programmes in a more ecient, safe and cost-eecve manner.

Informing Policy

This research has successfully demonstrated that water chemistry data collected using drone water sampling

methods are not stascally dierent from those produced by boat sampling. The studies undertaken have shown

that data precision and accuracy are not adversely impacted when using drone sampling compared with tradional

boat sampling methods. This comparave analysis sases the requirements of the European Union Water

Framework Direcve (WFD) sampling objecves for lake water monitoring and therefore can be applied to large-

scale water sampling programmes worldwide, such as the United Naons Global Environment Monitoring System for

Freshwater (GEMS/Water), Marine Strategy Framework Direcve and US Naonal Aquac Resource Surveys. Drone

water sampling also oers a unique opportunity to sample unmonitored lakes under the WFD in Ireland and remote

and inaccessible lakes worldwide, and conrm the water quality and ecological status of aquac environments

determined using remote sensing methods.

Developing Soluons

The project team is the rst research team in Ireland and Europe to capture a 2-L water sample using a drone. This

research has made several signicant contribuons towards the advancement and applicaon of drone water

sampling methods. These include the successful deployment, as demonstrated during eld trials, of a DJI M600 Pro

drone and aached payload capable of capturing a 2-L water sample and real-me physicochemical data, 100 metres

oshore, from open lakes in the west of Ireland. Water sampling mes using the drone were 4 minutes and water

volume capture rates were 100%. Drone water sampling was found to be 2.3 to 3.4 mes faster than boat sampling,

depending on resource allocaon. In contrast, however, the capital investment costs for boat sampling were found

to be 1.2 to 1.5 mes lower than those required for drone water sampling. However, drone water sampling reduced

the health and safety and biosecurity risks associated with boat sampling. Overall, the applicaon of drone water

sampling for large-scale water sampling programmes has been successfully demonstrated and can be adapted as

needed to aquac environments worldwide.

EPA Research Report 341

Assessing the Potenal of Drones to Take

Water Samples and Physico-chemical Data

from Open Lakes

Authors: Heather Lally, Ian O’Connor, Liam Broderick,

Mark Broderick, Olaf Jensen and Conor Graham

EPA Research: McCumiskey House,

Richiew, Clonskeagh, Dublin 14.

Phone: 01 268 0100

Twier: @EPAResearchNews

Email: research@epa.ie

EPA Research Webpages