نوع مقاله : مقاله پژوهشی
عنوان مقاله English
نویسندگان English
Pre-harvest crop yield prediction is crucial for food security, grain trade, and policy making. In this research, the use of random forest method in simulating maize yield in ten selected fields in Qazvin plain irrigation network during 2019-2020 period using NDVI, MSAVI, and EVI drought indices has been investigated. Sentinel 2 satellite was used for drought indices. The results were evaluated using R2, NRMSE, and MBE statistics. To investigate the relationship between drought indices and maize yield, seven scenarios were defined. The results showed that the random forest model in scenarios one, three, and four in the test stage with a significant probability of 95% respectively (P-value=0.00) and an explanation coefficient of more than 0.8 and a small amount of NRMSE is a suitable estimate of the maize yield.
کلیدواژهها English