Remotely Piloted Aircraft Systems (RPAS) application for mapping nitrogen deposition over intensively grazed grassland

JM Maire*, S Gibson-Poole, Nicholas Cowan, Karl G Richards, Ute M Skiba, RM Rees, Dave S Reay, Gary J Lanigan

*Corresponding author for this work

Research output: Chapter in Book/Report/Conference proceedingConference contribution

1 Downloads (Pure)

Abstract

In grazing livestock systems, the deposition of reactive nitrogen via urination is a significant potential source of nitrogen loses (e.g: nitrous oxide, ammonia emissions, nitrate leaching) which can account for 50 % to 60 % of the nitrogen input. These events occur randomly, resulting in high spatial variability at the field scale which prevents accurate accounting of their contribution to various pathways of nitrogen losses. This study investigated an alternative technique for identifying the spatial coverage in grasslands using Remotely Piloted Aircraft Systems (RPAS) technology. In this project, imagery from multi-rotor RPAS were used to identify urine patches in a 2 ha field (Johnstown Castle farm, Ireland) and in a 5 ha field (Easter Bush, Scotland), which had been grazed by dairy cows and sheep respectively. The field in Ireland was number of times surveyed over the grazing season of 2017, whilst the field in Scotland was surveyed once (in April 2016) to enable testing of the initial method. The results were summarised to construct a urine patch coverage map highlighting the size and colour properties of each patch. For the sheep grazing, the imagery of four samples of approximately 50 m2 areas within the field were analysed using a custom pixel based model written in R (the R Foundation, USA), that utilised colour channel thresholding and Kmeans clustering. For a total of 210 m2 of grassland, 4.12 % of the total area was considered influenced by urine events, with 82 patch areas averaging 0.11 m2. The detection of urine patches using RPAS imagery combined with soil measurements (greenhouse gas emissions, pH, moisture, nitrogen and carbon content) show potential to aid automatic and fast determination of urine patch cover at the field scale. This detection is essential for a better spatial modelling of nitrogen inputs, allowing better targeting of nitrogen fertilisers and the estimation of greenhouse gas emissions.
Original languageEnglish
Title of host publication20th EGU General Assembly, EGU2018, Proceedings from the conference held 4-13 April, 2018
Pages10520
Volume20
Publication statusPrint publication - 4 Apr 2018

Fingerprint

aircraft
grassland
farm
urine
nitrogen
imagery
grazing
sheep
greenhouse gas
nitrous oxide
targeting
livestock
pixel
ammonia
leaching
moisture
fertilizer
nitrate
carbon
modeling

Keywords

  • RPAS
  • UAV
  • Greenhouse
  • GHG
  • Urine

Cite this

Maire, JM., Gibson-Poole, S., Cowan, N., Richards, K. G., Skiba, U. M., Rees, RM., ... Lanigan, G. J. (2018). Remotely Piloted Aircraft Systems (RPAS) application for mapping nitrogen deposition over intensively grazed grassland. In 20th EGU General Assembly, EGU2018, Proceedings from the conference held 4-13 April, 2018 (Vol. 20, pp. 10520)
Maire, JM ; Gibson-Poole, S ; Cowan, Nicholas ; Richards, Karl G ; Skiba, Ute M ; Rees, RM ; Reay, Dave S ; Lanigan, Gary J. / Remotely Piloted Aircraft Systems (RPAS) application for mapping nitrogen deposition over intensively grazed grassland. 20th EGU General Assembly, EGU2018, Proceedings from the conference held 4-13 April, 2018. Vol. 20 2018. pp. 10520
@inproceedings{83e691cf00334613993af52330ba0260,
title = "Remotely Piloted Aircraft Systems (RPAS) application for mapping nitrogen deposition over intensively grazed grassland",
abstract = "In grazing livestock systems, the deposition of reactive nitrogen via urination is a significant potential source of nitrogen loses (e.g: nitrous oxide, ammonia emissions, nitrate leaching) which can account for 50 {\%} to 60 {\%} of the nitrogen input. These events occur randomly, resulting in high spatial variability at the field scale which prevents accurate accounting of their contribution to various pathways of nitrogen losses. This study investigated an alternative technique for identifying the spatial coverage in grasslands using Remotely Piloted Aircraft Systems (RPAS) technology. In this project, imagery from multi-rotor RPAS were used to identify urine patches in a 2 ha field (Johnstown Castle farm, Ireland) and in a 5 ha field (Easter Bush, Scotland), which had been grazed by dairy cows and sheep respectively. The field in Ireland was number of times surveyed over the grazing season of 2017, whilst the field in Scotland was surveyed once (in April 2016) to enable testing of the initial method. The results were summarised to construct a urine patch coverage map highlighting the size and colour properties of each patch. For the sheep grazing, the imagery of four samples of approximately 50 m2 areas within the field were analysed using a custom pixel based model written in R (the R Foundation, USA), that utilised colour channel thresholding and Kmeans clustering. For a total of 210 m2 of grassland, 4.12 {\%} of the total area was considered influenced by urine events, with 82 patch areas averaging 0.11 m2. The detection of urine patches using RPAS imagery combined with soil measurements (greenhouse gas emissions, pH, moisture, nitrogen and carbon content) show potential to aid automatic and fast determination of urine patch cover at the field scale. This detection is essential for a better spatial modelling of nitrogen inputs, allowing better targeting of nitrogen fertilisers and the estimation of greenhouse gas emissions.",
keywords = "RPAS, UAV, Greenhouse, GHG, Urine",
author = "JM Maire and S Gibson-Poole and Nicholas Cowan and Richards, {Karl G} and Skiba, {Ute M} and RM Rees and Reay, {Dave S} and Lanigan, {Gary J}",
year = "2018",
month = "4",
day = "4",
language = "English",
volume = "20",
pages = "10520",
booktitle = "20th EGU General Assembly, EGU2018, Proceedings from the conference held 4-13 April, 2018",

}

Maire, JM, Gibson-Poole, S, Cowan, N, Richards, KG, Skiba, UM, Rees, RM, Reay, DS & Lanigan, GJ 2018, Remotely Piloted Aircraft Systems (RPAS) application for mapping nitrogen deposition over intensively grazed grassland. in 20th EGU General Assembly, EGU2018, Proceedings from the conference held 4-13 April, 2018. vol. 20, pp. 10520.

Remotely Piloted Aircraft Systems (RPAS) application for mapping nitrogen deposition over intensively grazed grassland. / Maire, JM; Gibson-Poole, S; Cowan, Nicholas; Richards, Karl G; Skiba, Ute M; Rees, RM; Reay, Dave S; Lanigan, Gary J.

20th EGU General Assembly, EGU2018, Proceedings from the conference held 4-13 April, 2018. Vol. 20 2018. p. 10520.

Research output: Chapter in Book/Report/Conference proceedingConference contribution

TY - GEN

T1 - Remotely Piloted Aircraft Systems (RPAS) application for mapping nitrogen deposition over intensively grazed grassland

AU - Maire, JM

AU - Gibson-Poole, S

AU - Cowan, Nicholas

AU - Richards, Karl G

AU - Skiba, Ute M

AU - Rees, RM

AU - Reay, Dave S

AU - Lanigan, Gary J

PY - 2018/4/4

Y1 - 2018/4/4

N2 - In grazing livestock systems, the deposition of reactive nitrogen via urination is a significant potential source of nitrogen loses (e.g: nitrous oxide, ammonia emissions, nitrate leaching) which can account for 50 % to 60 % of the nitrogen input. These events occur randomly, resulting in high spatial variability at the field scale which prevents accurate accounting of their contribution to various pathways of nitrogen losses. This study investigated an alternative technique for identifying the spatial coverage in grasslands using Remotely Piloted Aircraft Systems (RPAS) technology. In this project, imagery from multi-rotor RPAS were used to identify urine patches in a 2 ha field (Johnstown Castle farm, Ireland) and in a 5 ha field (Easter Bush, Scotland), which had been grazed by dairy cows and sheep respectively. The field in Ireland was number of times surveyed over the grazing season of 2017, whilst the field in Scotland was surveyed once (in April 2016) to enable testing of the initial method. The results were summarised to construct a urine patch coverage map highlighting the size and colour properties of each patch. For the sheep grazing, the imagery of four samples of approximately 50 m2 areas within the field were analysed using a custom pixel based model written in R (the R Foundation, USA), that utilised colour channel thresholding and Kmeans clustering. For a total of 210 m2 of grassland, 4.12 % of the total area was considered influenced by urine events, with 82 patch areas averaging 0.11 m2. The detection of urine patches using RPAS imagery combined with soil measurements (greenhouse gas emissions, pH, moisture, nitrogen and carbon content) show potential to aid automatic and fast determination of urine patch cover at the field scale. This detection is essential for a better spatial modelling of nitrogen inputs, allowing better targeting of nitrogen fertilisers and the estimation of greenhouse gas emissions.

AB - In grazing livestock systems, the deposition of reactive nitrogen via urination is a significant potential source of nitrogen loses (e.g: nitrous oxide, ammonia emissions, nitrate leaching) which can account for 50 % to 60 % of the nitrogen input. These events occur randomly, resulting in high spatial variability at the field scale which prevents accurate accounting of their contribution to various pathways of nitrogen losses. This study investigated an alternative technique for identifying the spatial coverage in grasslands using Remotely Piloted Aircraft Systems (RPAS) technology. In this project, imagery from multi-rotor RPAS were used to identify urine patches in a 2 ha field (Johnstown Castle farm, Ireland) and in a 5 ha field (Easter Bush, Scotland), which had been grazed by dairy cows and sheep respectively. The field in Ireland was number of times surveyed over the grazing season of 2017, whilst the field in Scotland was surveyed once (in April 2016) to enable testing of the initial method. The results were summarised to construct a urine patch coverage map highlighting the size and colour properties of each patch. For the sheep grazing, the imagery of four samples of approximately 50 m2 areas within the field were analysed using a custom pixel based model written in R (the R Foundation, USA), that utilised colour channel thresholding and Kmeans clustering. For a total of 210 m2 of grassland, 4.12 % of the total area was considered influenced by urine events, with 82 patch areas averaging 0.11 m2. The detection of urine patches using RPAS imagery combined with soil measurements (greenhouse gas emissions, pH, moisture, nitrogen and carbon content) show potential to aid automatic and fast determination of urine patch cover at the field scale. This detection is essential for a better spatial modelling of nitrogen inputs, allowing better targeting of nitrogen fertilisers and the estimation of greenhouse gas emissions.

KW - RPAS

KW - UAV

KW - Greenhouse

KW - GHG

KW - Urine

M3 - Conference contribution

VL - 20

SP - 10520

BT - 20th EGU General Assembly, EGU2018, Proceedings from the conference held 4-13 April, 2018

ER -

Maire JM, Gibson-Poole S, Cowan N, Richards KG, Skiba UM, Rees RM et al. Remotely Piloted Aircraft Systems (RPAS) application for mapping nitrogen deposition over intensively grazed grassland. In 20th EGU General Assembly, EGU2018, Proceedings from the conference held 4-13 April, 2018. Vol. 20. 2018. p. 10520