Comparison of Unmanned Aerial Vehicle and Landsat Imagery to Monitor the Growth of Chickpea Crop in a Dryland Area of Thal Desert, Pakistan

Abstract

Remotely sensed data play an important role in providing meaningful information regarding crop management. Freely available remote sensing products (LandSat, Sentineil) has been used extensively; however, its low temporal, spectral, and spatial resolutions may limit its use. This study compared unmanned aerial vehicle (uav) and Landsat imagery in monitoring the crop growth parameters throughout the growing season in dryland agriculture system of Thal, Pakistan. study's main objective was to monitor and compare chickpea crop using landsat and uav imagery and crop yield estimation. In this study, correlation and regression analysis was performed between meteorological parametrs and yield Various vegetation indices i.e. normalized difference vegetation index (ndvi) and soil adjusted vegetation index (SAVI) were derived from uav and landsat imageries and compared with crop growth parameters and estimated final yield of chickpea crop. A coefficient (r2= 0.67; p≤0.05) between observed chickpea yield and uav derived savi was significant (p≤0.05). In contrast, a non-significant coefficient (r2=0.21; p≥0.05) was found between ndvi derived from Landsat imagery, and uav imagery vs yield. Early assessment of chickpea crop has been made by estimating chickpea crop yield 30-40 days before harvesting period using simple linear regression between ndvi and actual yield. The average difference between actual yield and predicted yield through savi-D and nadvi-L has been 558.42, 553.12, 556.84 kg/ha observed, respectively. This is the pioneer study for chickpea crops using uav (remote sensing technique), as drones provide quick use and easy access for small sections of crops with different combinations of sensors. The study suggests that crop monitoring could be assessed more accurately from uav remotely sensed data due to its high spatial resolution (<1m) and red-edge band for the studing the vegetation more accurately.