Maedeh Behifar; Mohsen Azadbakht; Farzaneh Hadadi; AliAkbar Matkan
Abstract
Vegetation indices are used to estimate vegetation parameters from satellite images. Despite their capabilities, performance of some vegetation indices decreases in high vegetation densities, making them inappropriate for estimation of the desired parameters. Vegetation indices are saturated in alfalfa ...
Read More
Vegetation indices are used to estimate vegetation parameters from satellite images. Despite their capabilities, performance of some vegetation indices decreases in high vegetation densities, making them inappropriate for estimation of the desired parameters. Vegetation indices are saturated in alfalfa farms due to the high chlorophyll content and high vegetation density; therefore, monitoring the changes of this plant is hindered. However, all indices do not perform similarly. In this research, the performance of different vegetation indices at different LAI values were investigated. The results showed that the CIgreen, CIrededge and NGRDI indices gained the best performance at high LAI values and they were less saturated. In contrast, the NDVI, NDREI and GNDI indices did not perform well and they were saturated at medium and high levels of LAI.
Farzaneh hadadi; Mohsen m_azadbakht; Maedeh Behifar; Hamid Salehi Shahrabi; amir moeinirad
Volume 10, Issue 3 , January 2019, , Pages 53-76
Abstract
Over the past several decades, many vegetation indices have been developed for crop yield estimation, each being sensitive to different levels of crop density and leaf area index, based on the bands and the algebraic formulas used in its design. However, the study of some perennial crops such as alfalfa, ...
Read More
Over the past several decades, many vegetation indices have been developed for crop yield estimation, each being sensitive to different levels of crop density and leaf area index, based on the bands and the algebraic formulas used in its design. However, the study of some perennial crops such as alfalfa, which are harvested several times annually, is very complicated and has received less attention. Therefore, in this paper, the most important vegetation indices developed to estimate alfalfa yield are using Sentinel-2 time series images. In this research, 144 alfalfa samples were collected periodically in a destructive way from alfalfa farms of Magsal Agricultural and Production Company (Qazvin) near the time of satellite pass, and then the efficiency of 10 of the most famous vegetation indices to estimate alfalfa yield was evaluated based on Sentinel-2 images. The results of this research showed that the estimated alfalfa yield using the index had the highest correlation () and the lowest root-mean-square-error (RMSE = 0.316 ) compared to the field data collected in the middle of August. In addition, the results showed that the red edge indices did not solve the saturation problem of vegetation indices and that the green vegetation indices were more capable of estimating alfalfa yield than the red edge indices.