作 者 ：Guo Liang, Du Peng, Xiao Qian-guang and Chen Wei-ying
Keywords:Remote Sensing, Transect, Evapotranspiration, Normalized difference vegetation index,
Abstract:The balance and exchange of water and energy in the ecosystem of China is noticeably affected by the monsoon climate especially associated with the occurrence of flood and drought in summer time. The determination of evapotranspiration (Ep), an important parameter of water balance, thus, is a linchpin for the investigation of monsoon-its mechanism and practical impacts. A new method to obtain actual Ep value has leen developed, which could monitor the actual Ep value in macro scale by using NOAA/AVHRR satallite data. The principle of the method is based on the experiment carried out by NASA in the eighth decade. In the IGBP( International Geosphere-Bio- sphere Program) studies, 13 special areas were selected for accessing their bio-diversity and the sensitiveness of climate change. NECT (Northeast China Transect) is one of the three mid-scale study areas. The location is 43.5˚ ± 1.5˚ N and 112˚ to 130.5˚ E. There are several biological systems including desert, dry grassland, steppe, forest and cultivated land. The vegetation index, as one of the significant indices of great importance in remote sensing, provides abundant and objective information for regional and global eco-environmental monitoring. NDVI (normalized difference vegetation index) is the most widely used vegetation index due to its efficiency, and is defined as: NDVI = ( CH2- CH1 )/( CH2 + CH1 ), where, CHI, CH2 are the albedo of channel 1 and channel 2 respectively. Some studies showed that NDVI is closely related to water supply conditions and Ep is one of the signs of water supply conditions as known, so some of our experiments was carried out to find the relation between Ep and NDVI in NECT. The experimental model of Ep is shown as following: E = a·e b·NDVI, where, a and b are the empirical coefficients. NDVI data of NECT specified in the years of 1990, 1991 and 1992 were selected in this study. The results are highlighted as follows: 1. The climate condition was normal in 1990; precipitation was abnormally high in 1991 as floods occurred in many places of China and droughts occurred in 1992 due to poor precipitation. The climate features of those years could be seen in the Ep map. For example, the area with Ep values from 0 to 250 mm was the smallest in 1991 and biggest in 1992, indicating that Ep reflects climate changes directly. 2. The Ep of China exhibited ribbon-like distribution, with much more prominent variation from west to east than from north to south. By analyzing the Ep in three lines (42N, 44N, 45N) of the transect, it was found that there was quite a similar trend of change in the lines from 200 mm to 800 nun or more. The distribution of the biological systems from the west to the east was as in the order of desert, dry grassland, steppe, dry grassland, cultivated land and forest. The patterns of Ep distribution were in agreement with the distribution of vegetation species in this area especially in the west of the transect. 3. Analysis of the inter-annual variation of Ep among the years of 1990, 1991 and 1992 revealed that one could deduct from the maximum variation (the absolute value) that, under a specific condition of climate change, the most significant response appeared to be at the boundaries between the dry grassland and the steppe where the vegetation population was very frangible and very sensitive to the slightest climate change. The result may be helpful in selecting a target area for further research on the effect and mechanism of monsoon climate.
全 文 ：