ESTIMATION OF LEAF AREA INDEX(LAI) OF POYANG LAKE BASIN IN SUMMER USING GF-1 DATA

GF-1 is a China's new high-resolution remote sensing satellite, which is characterized by high temporal and spatial resolutions as well as wide coverage. It provides new satellite resource to accurately and effectively monitor leaf area index (LAI). In this study, the monitoring of LAI was conducted with different land cover types during two times (July 26th July 30th, 2014, and July 28th August 5th, 2015), with the adoption of the Poyang Lake basin as the study area. Then the linear and nonlinear correlation analyses were conducted on the measured LAI data and the corresponding single-band data of GF-1 as well as 8 vegetation indexes (VIs). By constructing the LAI remote sensing inversion model, the high-accuracy LAI data were extracted. The research results indicate that, the data in the BLUE, GREEN, RED bands of GF-1 have high correlations with each other (R>0.89), but the data in near-infrared (NIR) band of GF-1 have poor correlations with the data in other three bands (R<0.3). The NIR and RED bands has strong reflection and absorption capacities separately, which is connected with the plant's spectral characteristics. LAI is negatively correlated with the reflectivity data in the BLUE, GREEN, RED bands of GF-1 but is positively correlated with VIs and the reflectivity data in the NIR band of GF-1. Most VIs are superior to single-band models in terms of the inversion of LAI. However, using the exponential model, the correlation between NDVI and the measured LAI is largest (R-2=0.744 and RMSE = 0.303). The results satisfy F test and have significant correlations. Accordingly, GF-1 satellite data is applicable to the monitoring of LAI with a high resolution over large areas.