Landsat Spectral Responses to Grassland Biophysical Conditions Across a Gradient in Inner Mongolia, China

Landsat;Biomass;Biodiversity;Natural Resources and Environment

Dai, JieBergen, Kathleen ;

University of Michigan

关键词: Landsat;    Biomass;    Biodiversity;    Natural Resources and Environment;   

Others  :  https://deepblue.lib.umich.edu/bitstream/handle/2027.42/99544/Jie_Dai_thesis_final%202013.pdf?sequence=1&isAllowed=y

瑞士|英语

来源: The Illinois Digital Environment for Access to Learning and Scholarship

【 摘 要 】

To investigate the potential of using Landsat imagery to detect grassland biophysical conditions, in particular biomass and biodiversity, harvested aboveground biomass and biodiversity were recorded along an ecological gradient in Inner Mongolia Autonomous Region (IMAR), China. Simultaneously vegetation spectral signatures were recorded by an analytical spectral device (ASD) Fieldspec 3 spectrometer. Vegetation indices (VIs) were calculated from the field spectrometer data following the same method as that of traditional Landsat-derived indices. Spatial regression analysis was used to assess the relationships between biomass and biodiversity and VIs. Based on maximum log likelihood and Akaike’s Information Criterion (AIC), we determined that the spatial error model between the log-transformations of both fresh biomass (lnBiom_f) and RVI (lnRVI) (R2=0.795, log = -13.77, AIC = 31.54) performed best in predicting fresh biomass for all sites. And the spatial error model between the log-transformations of both biodiversity (lnBiod) and RVI (lnRVI) (R2=0.763, log = -0.70, AIC = 5.40) performed best in predicting biodiversity through the ecological gradients in the entire study area. When predicting dry biomass, the spatial error model between the log-transformations of both dry biomass (lnBiom_d) and RVI (lnRVI) (R2=0.662, log = -20.28, AIC = 44.55) was the best, but the estimations for dry biomass were generally poor. This study verifies that Landsat data can reasonably monitor grassland biophysical conditions across large areas and different ecoregions.