Spatial upscaling of green aboveground biomass derived from MODIS-based NDVI in arid and semiarid grasslands

• Published in: Advances in space research Vol. 60; no. 9; pp. 2001 - 2008
• Main Authors: Xue, Juanjuan, Ge, Yonghui, Ren, Hongrui
• Format: Journal Article
• Language: English
• Published: Elsevier Ltd 01.11.2017
• Subjects: Arid and semiarid grasslands; Green aboveground biomass; Normalized difference vegetation index; Spatial upscaling; Green aboveground biomass; Normalized difference vegetation index; Spatial upscaling; Arid and semiarid grasslands
• ISSN: 0273-1177; 1879-1948
• DOI: 10.1016/j.asr.2017.07.016

Accurate estimation of green aboveground biomass is important for sustainable use of grassland resources in arid and semiarid grasslands. Nevertheless, it is difficult to achieve spatial upscaling of green aboveground biomass estimation using traditional spatial upscaling methods in arid and semiarid grasslands due to its inherent heterogeneity. In the study, a new spatial upscaling algorithm was proposed to estimate green aboveground biomass in the desert steppe of Inner Mongolia. The algorithm was successfully employed for spatial upscaling of green aboveground biomass estimation from MOD13Q1 NDVI (fine resolution) to MOD13A2 NDVI (coarse resolution) based on field measurements in the desert steppe. Results showed that, the correlation between distributed green aboveground biomass (obtained from fine resolution) and lumped green aboveground biomass (obtained from coarse resolution) was improved, and root mean squared error and relative error decreased after upscaling. Statistical analyses performing on the slopes and intercepts of the fitted lines between distributed green aboveground biomass and lumped green aboveground biomass demonstrated that, there was no significant difference (P>0.05) between the fitted line and the 1:1 line after upscaling, and there was significant difference (P<0.05) between the fitted line and the 1:1 line before upscaling. These indicated that, lumped green aboveground biomass after upscaling was much closer to distributed aboveground green biomass than lumped green aboveground biomass before upscaling. The algorithm proposed in the study could play an important role in large-scale green aboveground biomass investigation in arid and semiarid grasslands.