关键词: 藏北, 草地地上生物量, 遥感监测模型, 西藏高原

Abstract: Estimation of aboveground biomass (AGB) is necessary for studying productivity, carbon cycles, nutrient allocation, and fuel accumulation in terrestrial ecosystems. Remote sensing techniques allow scientists to examine properties and processes of ecosystems and their inter-annual variability at multiple scales since satellite observations can be obtained over large areas of interest with high re-visiting frequencies. In this study, AGB and its spatial distribution in the North Tibet are analyzed using in-situ measurements during August to September in 2004, and the regression models are established using in-situ AGB data and corresponding Terra MODIS NDVI and EVI. The results show that due to constraints of alpine climate, soil and moisture conditions, the average AGB in the North Tibet is very low, being 96.88 g/m². Of the total AGB, the dry matter content of fresh grass (hereafter referred to as fresh AGB) is above 80% and the dry matter content of dead material is below 20% . AGB raries greatly at different sampling sites and for different grassland types, ranging from 37.1-589.12 g/m². On an average, the largest AGB occurs in swampy meadow, being 356.84 g/m², followed by temperate steppe (64.48 g/m²) and alpine meadow (61.61 g/m²); the lowest value occurs in alpine steppe (48.87 g/m²). The optimum regression models for the North Tibet by using MODIS NDVI and EVI are compound, growth, exponent and power models, which can be used to estimate AGB and fresh AGB in the study area and four models have same performance. The efficiency of regression models based on MODIS NDVI or EVI for fresh AGB is better than for total AGB, which reflects unique spectral response of green plant. At spatial level, both AGB and fresh AGB decrease from the southeast to the northwest in the study area, reaching above 100 g/m² in some part of southeastern region while below 20 g/m² in northwestern region.

Key words: aboveground biomass, remote sensing model, North Tibet, grassland