Masoud Eshghizadeh; Yaser Esmaeilian
Volume 27, Issue 1 , April 2020, , Pages 159-176
Abstract
Due to the limitations of field measurements of vegetation, the application of plant indexes to estimate rangeland biomass using satellite data can be very useful in rangeland studies. For this purpose, it is necessary to identify appropriate vegetation indices. The aim of this study is to investigate ...
Read More
Due to the limitations of field measurements of vegetation, the application of plant indexes to estimate rangeland biomass using satellite data can be very useful in rangeland studies. For this purpose, it is necessary to identify appropriate vegetation indices. The aim of this study is to investigate the possibility of estimating rangeland biomass using plant indices obtained from digital data of Landsat 8 satellite and determining the most appropriate ones in semi-arid regions of the northeast of the country. For this purpose, the average values of plant indices NDVI, TDVI, SAVI, ARVI, EVI, OSAVI, IPVI, GRVI, and GNDVI within each unit of one hectare of the studied basin network were calculated. Then, the correlation of these values with the average measured field values of these units was examined by linear regression, and the regression model of each index was determined to estimate biomass. Finally, the results were validated and a field biomass map was prepared for each index. The results showed that all indexes had a high and acceptable correlation with real biomass data. Based on the validation results, the SAVI plant index with a coefficient 0.79 and root-mean-square error of 14.73% was the most suitable plant index for estimating biomass in the region. By using the wavelengths located in the blue band, these indicators modify the effect of dust in the calculations, which reduces the atmospheric effect and improves the results of calculating the NDVI index, and it can be called the modified NDVI index. According to the results, plant indices obtained from the ratio of near and visible infrared bands are highly correlated with biomass. In general, the shorter the wavelengths used by plant indices, the lower the accuracy of estimates in arid and semi-arid regions.