Effects of spatial, temporal, and spectral resolutions on the estimation of wheat and barley leaf area index using multi- and hyper-spectral data (case study: Karaj, Iran)

Leaf area index (LAI) is a key parameter for the calculation of crop biophysical and biochemical processes. Therefore, the accurate estimates of LAI has been always of great importance for agricultural researchers. Remote sensing has shown enormous potential in LAI estimation, however, more evaluations are necessary on choosing the best type of data. In this study, the spatial, temporal, and spectral resolutions of different remotely sensed data (Landsat 8, Sentinel-2, MODIS, and also field hyperspectral data) were evaluated for LAI estimation of wheat and barley. First, the 30-m Landsat 8, 10-m Sentinel-2, 250-m MODIS, and field-based point data were taken into account for assessing the goodness of the relationship between field LAI (collected using LAI-2200c) and Vegetation Indices (VIs) to investigate the effect of a difference in spatial resolution. Afterward, to assess the temporal resolution effects, the Sentinel-2 images were resampled to 30 m and were combined with Landsat 8 data. Also, hyperspectral VIs (HNDVI, HDVI, and HSR) were calculated using field data to evaluate the effects of spectral resolution. Results showed that the difference in spatial and temporal resolutions of the data did not have any considerable effect on improving the LAI-VI relationship. Nevertheless, there were some particular portions of the spectrum which had R2 of more than 0.8 which was a great improvement compared to multispectral data with R2 between 0.6 and 0.69. The best HNDVI and HSR were calculated from the 10-nm bands centered at 1 115 nm and 1 135 nm.