Regional Symposium on Innovation in science and technology

فارسی

Home / دومین سمپوزیوم منطقه ای نوآوری در علم و فناوری

Monitoring Wheat Crop Phenology in semi-arid agroclimatic region using UAV-RGB and high-resolution multispectral imaging system

Authors :

Sana Arshad¹ Jamil Kazmi² Saima Shaikh³ Maryam Khan⁴

1- Department of Geography, the Islamia University of Bahawalpur 2- Department of Geography, University of Karachi, Karachi 75270, Pakistan 3- Department of Geography, University of Karachi, Pakistan 4- Department of Geography, University of Karachi, Pakistan

Keywords :

DJI Mavic،NDVI،VARI،heat stress

Abstract :

Accurate monitoring of Wheat crop phenology is substantial to understand the impact of climate change on its production and yield. For this purpose, several techniques of precision farming are being adopted nowadays at farm level agricultural management. The utilization of unnamed aerial vehicles (UAV) or drones is quite significant in this regard in varied climatic conditions. In this context, our study aimed to identify and monitor the wheat crop phenology employing UAV application in an open experimental site of district Bahawalpur, Pakistan. The Quadcopter Drone i.e., DJI Mavic Air 2s (RGB FC3411-camera) with 1´´ CMOS sensor was used in the research. Six UAV-flights were planned at an altitude of 55meters on a 20-acre experimental site following the wheat crop calendar at different stages of crop growth in the sequence of 5th November (sowing-leaf development stage), 1st December (tillering-Booting), 15th January (flowering-fruit development), 20th February (ripening), 14th March (grain filled), 27th March (before harvest). More than 400 vertical georeferenced UAV images are stitched to make the ortho-mosaic in Pix4D and ArcGIS Pro with three major outputs: ortho-mosaic (OM), digital surface model (DSM), and point cloud. The ortho-mosaic images were used for further analyzing the RBG Index based crop phenological identification and mapping. The Visible Atmospherically Resistant Index (VARI) was computed from red, green, and blue channel of each ortho-image in ArcGIS-pro. The developed time series of VARI in respective wheat season was synchronized with multispectral Sentinel-2 NDVI. Moreover, environmental characteristics (maximum temperature, minimum temperature, rainfall, relative humidity, wind speed, and heat stress) were also considered to analyze with wheat-VARI and wheat-NDVI. Results revealed that VARI accurately identified the wheat phenology in the study area with a significant (p<0.001) high correlation (r = 0.77) with NDVI. Moreover, less precipitation was observed during the whole wheat growth cycle with minimum and zero correlation with VARI. Heat stress was observed at the tillering (November) and grain-filling to ripening (March-Apr) stage of wheat growth with negative correlation (r = -0.33) with VARI and NDVI. Hence, study provides an effective utilization of UAV based precision farming technology for crop growth and its relationship with environmental traits.