Estimation Of Forest Net Primary Production Of Jing-Jin-Ji Region Based On Remote Sensing

Carbon cycling of forest ecosystem is a crucial issue both for global climate changes and global carbon cycling. The estimation of vegetation net primary production (NPP) is the basis and important content of global carbon cycling research. Based on previous results, this thesis improved Carnegie Ames Stanford Approach (CASA) model and developed the estimation software under VC++ programing environment. Supported by this software, remote sensing and GIS (geographic information system ) technology, NPP of Jing-Jin-Ji region in 2004 was retrieved using vegetation, meteorological and soil data, and the temporal and spatial distribution patterns of NPP were analyzed. Also the interactions between the temporal and spatial distribution patterns of NPP and climate factors such as air temperature and precipitation were discussed.The main conclusions are following:1. The CASA model improvementWith regard to the differences of maximum utilization efficiency in different vegetation types and the actual situation of China's vegetation, an improved CASA model, based on the basic framework of CASA, was presented. Through the verification of the NPP simulated results, it showed that the improved CASA model has better effect on simulating vegetation NPP at regional scale.2. The estimation of forest vegetation NPP of Jing-Jin-Ji region based on remote sensing Based on the improved CASA model, the estimation of forest vegetation NPP of Jing-Jin-Ji region in 2004 was realized using MODIS data (spatial resolution is 1Km, temporal resolution is 8 days). The results indicated that the total forest vegetation NPP of study area is 14884.18 KtC/ a, average NPP is 658.8 gC / m2?a, and varied value is from zero to 1200 gC / m2?a, mainly concentrated range of 300～800 gC / m2?a. At the same time, APAR andεof study area was also estimated in this study.3. The temporal and spatial distribution patterns of NPP and its relationship with climatic factors Through analyzing spatial and temporal distribution patterns of NPP, we found that forest vegetation NPP were mainly distributed in the Yanshan mountain region, Taihang Mountains region have banded distribution, and were scattered in other regions; The forest vegetation NPP seasonal changes are obvious, showing a single peak curve, which mainly accumulated during from May to September and at other times is lower; Through analyzing the relationship between NPP spatial and temporal distribution patterns and climatic factors, it is concluded that temperature and precipitation is main factors influenced temporal and spatial distribution of forest vegetation NPP of the study area. In addition, the effect of temperature is stronger than that of precipitation. 4. Through analyzing the seasonal changes of NPP, we proposed a general idea that using high temporal resolution images accrurately estimate NPP at microscale based on high spatial resolution images.