Agricultural Water Management (Sep 2023)

Evaluation of UAV-based drought indices for crop water conditions monitoring: A case study of summer maize

  • Minghan Cheng,
  • Chengming Sun,
  • Chenwei Nie,
  • Shuaibing Liu,
  • Xun Yu,
  • Yi Bai,
  • Yadong Liu,
  • Lin Meng,
  • Xiao Jia,
  • Yuan Liu,
  • Lili Zhou,
  • Fei Nan,
  • Tengyu Cui,
  • Xiuliang Jin

Journal volume & issue
Vol. 287
p. 108442

Abstract

Read online

Accurately monitoring the crop water conditions (CWC) is vital for agricultural water management. Traditional in situ measurements are limited by inefficiency and lack of spatial information. However, the development of unmanned aerial vehicle (UAV) applications in agriculture now provides a high throughput and cost-effective method to obtain field crop growth information. Unfortunately, current UAV-based drought indices do not capture the time series information, or the accuracy is limited. This study uses UAV-based multispectral and thermal information and site-observed air temperature to obtain the following three UAV-based drought indices: the normalized relative canopy temperature (NRCT), the temperature vegetation drought index (TVDI), and the three-dimension drought index (TDDI). We evaluate the accuracy with which these indices can be used to characterize the CWC of field maize by comparing them with in situ vegetation moisture contents (VMC). This study aims to (i) evaluate the pertinence of the TDDI for characterizing VMC, (ii) compare the performance of TDDI with that of NRCT and TVDI, and (ii) analyze the spatiotemporal variation of the three drought indices. The results show that (i) TDDI provides the best estimates of VMC (r = 0.71), (ii) NRCT and TVDI are comparable for characterizing VMC (r = 0.59 and 0.63, respectively) and are strongly correlated (r = 0.92), (iii) the three indices characterize the spatial distribution of VMC well, but the multi-phase image information used by TDDI makes it significantly better for studying VMC temporal variations than NRCT and TVDI. The results of this study prove that UAV-based observations can be used to accurately monitor field crop water conditions. In addition, the TDDI provides new insights into the study of remote sensing-based drought indices.

Keywords