Prediction of leaf area index using integration of the thermal infrared and optical data over the mixed temperate forest

Although the retrieval of leaf area index (LAI) as one of the essential biodiversity variable from remote sensing data has shown to be successful over visible/near-infrared (VNIR, 0.3-1.0 μm), shortwave infrared (SWIR, 1.0-2.5 μm), and TIR (8-14 μm) domains, integration of VNIR/SWIR with the TIR data for LAI estimation has not been addressed yet. Despite the importance, maturity, and availability of the remotely sensed data over VNIR and SWIR regions, TIR remote sensing data (i.e., emissivity spectra) has a number of advantages for LAI estimation. As such, it is known that the emissivity spectra over the TIR domain do not saturate even at relatively high values of LAI. In this respect, the utility of Landsat-8 TIR data together with its optical spectral data was examined to quantify LAI over Bavarian Forest National Park (Mixed temperate forest) in Germany. A field campaign was conducted in August 2015 in the National Park concurrent with the time of the Landsat-8 overpass. LAI was measured in the field for 37 plots. In this study, a number of vegetation indices, which have been widely applied in the literature were used to estimate LAI using VNIR/SWIR data. Furthermore, land surface emissivity (i.e., LSE) was derived from the band 10 of TIRS sensor using the normalized difference vegetation index threshold method. LSE was integrated with the reflectance data as the input layers to examine the LAI retrieval accuracy using the artificial neural network as a machine learning approach. The levenberg-marquardt algorithm was used for network training. LAI was predicted with modest accuracy using vegetation indices (R2CV=0.63, RMSECV=1.56 m2m-2, and R2CV=0.65, RMSECV=1.56 m2m-2 for NDI, and SR respectively). However, when the VNIR/SWIR bands and TIR data (LSE) were integrated, the prediction accuracy of LAI increased significantly (R2CV=0.79, RMSECV=0.75, m2m-2). Our results demonstrate that the combination of LSE and VNIR/SWIR satellite data can lead to higher retrieval accuracy for LAI. This finding has implication for retrieval of other vegetation parameters through the integration of TIR and optical satellite remote sensing data as well as regional mapping of LAI when coupled with a canopy radiative transfer model.
3