ABSTRACT
The topography effects on vegetation biomass under climate change impact have been ignored in prairie regions as it is not as significant as in mountain areas. This paper aims to investigate the topographic effects on vegetation biomass under climate change in semiarid Canadian mixed grass prairie. The study site is Grasslands National Park (GNP) and the study period is from 1985 to 2007. Data used include dry green biomass data sampled from June to July of 2003 to 2005, 10-day Advanced Very High Resolution Radiometer (AVHRR) 1km Normalized Difference Vegetation Index (NDVI) composites of 1985 to 2007, and Global Digital Elevation Model derived from Advanced Spaceborne Thermal Emission and Reflection Radiometer (ASTER GDEM) data with 90 m resolution. To achieve the objective, the applicability of AVHRR NDVI data being a proxy of vegetation biomass was investigated. Then, the range and standard deviation (SD) of each individual vegetation patch in both valley and upland grasslands were calculated. In addition, the variation trend of valley and upland vegetation was analyzed respectively using the Mann-Kendall (M-K) test and the Sen’s slope. The results indicate that the interannual variation of vegetation biomass at GNP can be fairly well represented by AVHRR 1 km NDVI data. Although some patches in valley grassland have similar NDVI range and SD values as those in upland grassland, the others have much smaller range and SD values than the highest range (0.154) and SD (0.045) of upland grassland. The M-K test and Sen’s slope analyses indicate that NDVI had an increase trend with a larger slope (0.0005) in upland and a smaller slope (0.0002) in valley grassland. It is concluded that climatic variation has more effects on upland grassland than valley grassland in GNP. Topography effects in prairie regions should not be ignored.