| As one of the most important types of terrestrial ecosystems on the planet, grassland ecosystem plays an important role in the global terrestrial circulation among ecosystems. Grassland productivity is the most direct indicator of grassland’s ecological status, thus, precise and rapid estimation of grassland productivity is the hot research area and important scientific issue. Specially, research on grassland Net Primary Productivity(NPP) has an important significance for evaluating regional ecological carrying capacity, utilizing and developing natural resources reasonably, and guaranteeing the ecological security and sustainable development of our country.In this study, based on MODIS remote sensing data, meteorological data and ground measured data, we modified the important parameters of CASA model, built a regional model of NPP aimed at Xilingol grassland, and valuated model accuracy. We also used the modified model to estimate NPP of Xilingol grassland during 2005 to 2014, and further analyzed its spatio-temporal distribution. The main results were follows:(1) According to the least error criterion between simulated values and measured values of NPP, we calculated the largest utilization efficiency of solar energy to be 0.539 g_C.MJ-1.(2) There existed significant correlation between simulated NPP and measured NPP, and the correlation coefficient was about 0.72. The mean relative estimation error(REE) between simulated values and the measured values was 0.29, this suggested that the accuracy of modified CASA model was 71% compared with measured NPP, which can meet accuracy requirement.(3) NPP of Xilingol grassland decreased from northeast to southwest gradually, the region with NPP great than 400 g_C.m-2.a-1 mainly distributed in northeast of Xilingol, NPP ranged 100 to 200 g_C.m-2.a-1 in the central region, and was generally less than 100 g_C.m-2.a-1 in the southwest region.(4) There was a significant difference in NPP of different grassland types, the overall trend of NPP was as follows: the eastern meadow grassland > the central typical grassland > the western desert grassland. Specially, annual average NPP per unit area were greater than the overall level of Xilingol grassland in montane meadow, temperate meadow-steppe and Low-land meadow; and less than the overall level of Xilingol grassland in temperate steppe, improved grassland, temperate desert-steppe, temperate desert and temperate steppe-desert.(5) The total annual NPP showed a rising trend during 2005 to 2014, increasing from 156.67 g_C.m-2.a-1 in 2005 to 186.53 g_C.m-2.a-1 in 2014. There was the maximum annual NPP in 2008, reaching 187.57 g_C.m-2.a-1, and 120 g_C.m-2.a-1 was the minimum annual NPP in 2008. In the same region, precipitation and temperature were the main factors to affect NPP in the short term, but the influence mechanism these two factors to NPP were not the same in different conditions. In terms of research area, when precipitation was sufficient, the rise of temperature didn’t have great influence on NPP; however, the rise of temperature would decrease NPP when precipitation was less.(6) There was a great difference in NPP variation coefficient of different grassland types during 2005 to 2014, meadow steppe and temperate steppe with high productivity changed smoothly in different years, while temperate desert-steppe and temperate steppe-desert with low productivity changed greatly in different years. The date with maximum NPP of different grassland types delayed by the order of meadows, steppe and desert. |
