Unmanned aerial vehicle-based assessment of cover crop biomass and nitrogen uptake variability

M. Yuan, J. C. Burjel, J. Isermann, N. J. Goeser, Cameron M Pittelkow

Research output: Contribution to journalArticle

Abstract

Cover crops have the potential to reduce nitrate (NO3) losses while improving soil quality, yet achieving uniform cover crop establishment can be challenging in the US Midwest. Understanding the variability of cover crop biomass and nitrogen (N) uptake at the field-scale is an important step in determining potential effects on the following cash crop and benefits to water quality, but efficient and nondestructive methods are lacking. The objective of this study was to evaluate a lightweight unmanned aerial vehicle (UAV) mounted with a multispectral sensor in estimating cover crop (grass species) biomass and N uptake prior to cover crop termination at four commercial fields in Illinois during the 2017 growing season. Two fields had triticale (× Triticosecale Wittmack) before corn (Zea mays L.) and two had cereal rye (Secale cereale L.) before soybean (Glycine max [L.] Merr.). Forty ground-truth biomass samples (1 m2) were collected on the day of each UAV flight across each field. Linear relationships were established between cover crop biomass and N uptake and four vegetation indices (VIs; Normalized Difference Vegetation Index [NDVI], Green Ratio Vegetation Index [GRVI], Green Normalized Difference Vegetation Index [GNDVI], and Triangular Vegetation Index [TVI]). The four VIs performed similarly in estimating cover crop biomass and N uptake (R2 range, 0.42 to 0.93; RMSE range, 9.4% to 27.2% of the range of biomass or N uptake). A high degree of within-field variability for NDVI was observed at all fields, with biomass and N uptake at soybean fields ranging from 0 to 1,790 kg ha-1 and 0 to 48.5 kg ha-1, respectively, and at corn fields 0 to 840 kg ha-1 and 0 to 31.5 kg ha-1, respectively. Because cover crop biomass is often estimated based on hand samples, we also simulated the effects of biomass sampling number (2, 5, 10, or 15) on the probability of reaching different accuracy levels for estimating field means for different field sizes. Under the growing conditions of this single study year and relatively modest biomass accumulation (<1,400 kg ha-1), results from this preliminary study provide evidence that UAVs are a viable technique to obtain relatively rapid, nondestructive estimates of biomass and N uptake of two grass cover crops at vegetative stage prior to termination at the field-scale in the US Midwest. This approach could help effectively utilize scarce conservation resources, but further work is needed to evaluate other cover crop species under a wider range of growth conditions.

Original languageEnglish (US)
Pages (from-to)350-359
Number of pages10
JournalJournal of Soil and Water Conservation
Volume74
Issue number4
DOIs
StatePublished - Jan 1 2019

Fingerprint

cover crop
cover crops
uptake mechanisms
nitrogen
biomass
vegetation index
NDVI
soybean
unmanned aerial vehicles
vehicle
soybeans
maize
grasses
grass
Triticosecale
cash crops
corn
nondestructive methods
plant establishment
Secale cereale

Keywords

  • Cereal rye
  • Cover crops
  • Normalized difference vegetation index (NDVI)
  • Remote sensing
  • Unmanned aerial vehicle (UAV)
  • Vegetation indices

ASJC Scopus subject areas

  • Agronomy and Crop Science
  • Water Science and Technology
  • Soil Science
  • Nature and Landscape Conservation

Cite this

Unmanned aerial vehicle-based assessment of cover crop biomass and nitrogen uptake variability. / Yuan, M.; Burjel, J. C.; Isermann, J.; Goeser, N. J.; Pittelkow, Cameron M.

In: Journal of Soil and Water Conservation, Vol. 74, No. 4, 01.01.2019, p. 350-359.

Research output: Contribution to journalArticle

@article{1401dfee56174c97b4be2a3b205382fc,
title = "Unmanned aerial vehicle-based assessment of cover crop biomass and nitrogen uptake variability",
abstract = "Cover crops have the potential to reduce nitrate (NO3) losses while improving soil quality, yet achieving uniform cover crop establishment can be challenging in the US Midwest. Understanding the variability of cover crop biomass and nitrogen (N) uptake at the field-scale is an important step in determining potential effects on the following cash crop and benefits to water quality, but efficient and nondestructive methods are lacking. The objective of this study was to evaluate a lightweight unmanned aerial vehicle (UAV) mounted with a multispectral sensor in estimating cover crop (grass species) biomass and N uptake prior to cover crop termination at four commercial fields in Illinois during the 2017 growing season. Two fields had triticale (× Triticosecale Wittmack) before corn (Zea mays L.) and two had cereal rye (Secale cereale L.) before soybean (Glycine max [L.] Merr.). Forty ground-truth biomass samples (1 m2) were collected on the day of each UAV flight across each field. Linear relationships were established between cover crop biomass and N uptake and four vegetation indices (VIs; Normalized Difference Vegetation Index [NDVI], Green Ratio Vegetation Index [GRVI], Green Normalized Difference Vegetation Index [GNDVI], and Triangular Vegetation Index [TVI]). The four VIs performed similarly in estimating cover crop biomass and N uptake (R2 range, 0.42 to 0.93; RMSE range, 9.4{\%} to 27.2{\%} of the range of biomass or N uptake). A high degree of within-field variability for NDVI was observed at all fields, with biomass and N uptake at soybean fields ranging from 0 to 1,790 kg ha-1 and 0 to 48.5 kg ha-1, respectively, and at corn fields 0 to 840 kg ha-1 and 0 to 31.5 kg ha-1, respectively. Because cover crop biomass is often estimated based on hand samples, we also simulated the effects of biomass sampling number (2, 5, 10, or 15) on the probability of reaching different accuracy levels for estimating field means for different field sizes. Under the growing conditions of this single study year and relatively modest biomass accumulation (<1,400 kg ha-1), results from this preliminary study provide evidence that UAVs are a viable technique to obtain relatively rapid, nondestructive estimates of biomass and N uptake of two grass cover crops at vegetative stage prior to termination at the field-scale in the US Midwest. This approach could help effectively utilize scarce conservation resources, but further work is needed to evaluate other cover crop species under a wider range of growth conditions.",
keywords = "Cereal rye, Cover crops, Normalized difference vegetation index (NDVI), Remote sensing, Unmanned aerial vehicle (UAV), Vegetation indices",
author = "M. Yuan and Burjel, {J. C.} and J. Isermann and Goeser, {N. J.} and Pittelkow, {Cameron M}",
year = "2019",
month = "1",
day = "1",
doi = "10.2489/jswc.74.4.350",
language = "English (US)",
volume = "74",
pages = "350--359",
journal = "Journal of Soils and Water Conservation",
issn = "0022-4561",
publisher = "Soil Conservation Society of America",
number = "4",

}

TY - JOUR

T1 - Unmanned aerial vehicle-based assessment of cover crop biomass and nitrogen uptake variability

AU - Yuan, M.

AU - Burjel, J. C.

AU - Isermann, J.

AU - Goeser, N. J.

AU - Pittelkow, Cameron M

PY - 2019/1/1

Y1 - 2019/1/1

N2 - Cover crops have the potential to reduce nitrate (NO3) losses while improving soil quality, yet achieving uniform cover crop establishment can be challenging in the US Midwest. Understanding the variability of cover crop biomass and nitrogen (N) uptake at the field-scale is an important step in determining potential effects on the following cash crop and benefits to water quality, but efficient and nondestructive methods are lacking. The objective of this study was to evaluate a lightweight unmanned aerial vehicle (UAV) mounted with a multispectral sensor in estimating cover crop (grass species) biomass and N uptake prior to cover crop termination at four commercial fields in Illinois during the 2017 growing season. Two fields had triticale (× Triticosecale Wittmack) before corn (Zea mays L.) and two had cereal rye (Secale cereale L.) before soybean (Glycine max [L.] Merr.). Forty ground-truth biomass samples (1 m2) were collected on the day of each UAV flight across each field. Linear relationships were established between cover crop biomass and N uptake and four vegetation indices (VIs; Normalized Difference Vegetation Index [NDVI], Green Ratio Vegetation Index [GRVI], Green Normalized Difference Vegetation Index [GNDVI], and Triangular Vegetation Index [TVI]). The four VIs performed similarly in estimating cover crop biomass and N uptake (R2 range, 0.42 to 0.93; RMSE range, 9.4% to 27.2% of the range of biomass or N uptake). A high degree of within-field variability for NDVI was observed at all fields, with biomass and N uptake at soybean fields ranging from 0 to 1,790 kg ha-1 and 0 to 48.5 kg ha-1, respectively, and at corn fields 0 to 840 kg ha-1 and 0 to 31.5 kg ha-1, respectively. Because cover crop biomass is often estimated based on hand samples, we also simulated the effects of biomass sampling number (2, 5, 10, or 15) on the probability of reaching different accuracy levels for estimating field means for different field sizes. Under the growing conditions of this single study year and relatively modest biomass accumulation (<1,400 kg ha-1), results from this preliminary study provide evidence that UAVs are a viable technique to obtain relatively rapid, nondestructive estimates of biomass and N uptake of two grass cover crops at vegetative stage prior to termination at the field-scale in the US Midwest. This approach could help effectively utilize scarce conservation resources, but further work is needed to evaluate other cover crop species under a wider range of growth conditions.

AB - Cover crops have the potential to reduce nitrate (NO3) losses while improving soil quality, yet achieving uniform cover crop establishment can be challenging in the US Midwest. Understanding the variability of cover crop biomass and nitrogen (N) uptake at the field-scale is an important step in determining potential effects on the following cash crop and benefits to water quality, but efficient and nondestructive methods are lacking. The objective of this study was to evaluate a lightweight unmanned aerial vehicle (UAV) mounted with a multispectral sensor in estimating cover crop (grass species) biomass and N uptake prior to cover crop termination at four commercial fields in Illinois during the 2017 growing season. Two fields had triticale (× Triticosecale Wittmack) before corn (Zea mays L.) and two had cereal rye (Secale cereale L.) before soybean (Glycine max [L.] Merr.). Forty ground-truth biomass samples (1 m2) were collected on the day of each UAV flight across each field. Linear relationships were established between cover crop biomass and N uptake and four vegetation indices (VIs; Normalized Difference Vegetation Index [NDVI], Green Ratio Vegetation Index [GRVI], Green Normalized Difference Vegetation Index [GNDVI], and Triangular Vegetation Index [TVI]). The four VIs performed similarly in estimating cover crop biomass and N uptake (R2 range, 0.42 to 0.93; RMSE range, 9.4% to 27.2% of the range of biomass or N uptake). A high degree of within-field variability for NDVI was observed at all fields, with biomass and N uptake at soybean fields ranging from 0 to 1,790 kg ha-1 and 0 to 48.5 kg ha-1, respectively, and at corn fields 0 to 840 kg ha-1 and 0 to 31.5 kg ha-1, respectively. Because cover crop biomass is often estimated based on hand samples, we also simulated the effects of biomass sampling number (2, 5, 10, or 15) on the probability of reaching different accuracy levels for estimating field means for different field sizes. Under the growing conditions of this single study year and relatively modest biomass accumulation (<1,400 kg ha-1), results from this preliminary study provide evidence that UAVs are a viable technique to obtain relatively rapid, nondestructive estimates of biomass and N uptake of two grass cover crops at vegetative stage prior to termination at the field-scale in the US Midwest. This approach could help effectively utilize scarce conservation resources, but further work is needed to evaluate other cover crop species under a wider range of growth conditions.

KW - Cereal rye

KW - Cover crops

KW - Normalized difference vegetation index (NDVI)

KW - Remote sensing

KW - Unmanned aerial vehicle (UAV)

KW - Vegetation indices

UR - http://www.scopus.com/inward/record.url?scp=85068740226&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85068740226&partnerID=8YFLogxK

U2 - 10.2489/jswc.74.4.350

DO - 10.2489/jswc.74.4.350

M3 - Article

AN - SCOPUS:85068740226

VL - 74

SP - 350

EP - 359

JO - Journal of Soils and Water Conservation

JF - Journal of Soils and Water Conservation

SN - 0022-4561

IS - 4

ER -