陕北黄土高原植被净初级生产力的估算

基于MODIS和地面气象数据,利用改进的CASA模型,模拟分析了2005年陕北黄土高原地区的植被净初级生产力(NPP)及其时空分布.结果表明:1)根据生态生理过程模型针对不同土地覆被类型选择不同的月平均最大光能利用率,比传统CASA模型中使用固定的全球月平均最大光能利用率进行NPP估算,更符合陕北黄土高原地区的实际情况;在估算植被参数时引入植被覆盖分类,以及利用陕北黄土高原2005年时序NDVI进行土地覆被分类的同时,结合1:100万中国植被图和实地调查情况对分类结果进行修正,可提高分类的精度,从而提高模型估算的精度.2)通过不同模型之间和与陕北部分地区实际调查数据进行比较,显示改进后的CASA模型对区域陆地植被NPP的模拟效果较好,可应用于陕北黄土高原乃至周边地区NPP的计算中.3)2005年陕北黄土高原植被净第一性生产量估计值为4.76×10~(13) g C,约占全国总NPP的1.5%,植被平均NPP为447.3 g C·m~(-2)·a~(-1),高于1992-2000年全国陆地NPP平均值323.8 g C·m~(-2)·a~(-1).4)在NPP的空间分布上,总体趋势是由东南向西北递减,其中最高值出现在东南部的黄龙山次生林区(1087g C·m~(-2)·a~(-1));西北部的荒漠植被覆盖度极低,平均NPP仅为205.0 g C·m~(-2)·a~(-1).5)陕北黄土高原NPP的季节变化明显,其中4月中旬至10月中旬6个月生长季时间里的NPP可占到全年的91.5%,而7月中旬至8月中旬间该区的净初级生产力达到年内的极大值,可占全年的37.8%.

Abstract：

Based on the data from MODIS (Moderate-resolution Imaging Speetroradiometer) and meteorological observatories, and by using improved CASA (Carnegie-Ames-Stanford Approach) model, the vegetation net primary productivity (NPP) and its spatiotemporal distribution on the North Shaanxi Loess Plateau in 2005 were simulated and analyzed. Comparing with the traditional CASA model which only uses a universal mean annual maximum light use efficiency (LUE), the estimated regional NPP by the improved CASA model was more precise, because this improved model used the LUE parameters of different vegetation covers. The detailed land cover classifica-tion also contributed to the increase of the precision via introducing the time-series Normalized Different Vegetation Index (NDVI) and ground survey data to modify and adjust the original clas-sification system based on vegetation map (1: 1000000). The testing of the simulation results from different models with the ground survey data in North Shaanxi showed that the estimation by the modified CASA model was much closer to the real survey data, implying the potential practi-cal significance of this model in estimating the vegetation NPP in North Shaanxi Loess Plateau and its adjacent areas. In 2005, the NPP in North Shaanxi was estimated as 4. 76×1013 g C, ac-counting for about 1.5% of China' s terrestrial total NPP, and the mean NPP was 447.3 g C·m~(-2)·a~(-1), being much higher than that of China' s terrestrial vegetation (323.8 g C·m~(-2)·a~(-1)) in 1982-2000. The spatial distribution pattern of the vegetation NPP showed an apparently declining trend from the southeast to the northwest, with the highest value of 1087 g C·m~(-2)a~(-1) occurred in the broadleaved-and conifer-mixed forests of Huanglong Mountain in southeast part of the region. The mean NPP of desert vegetation in the whole region was the lowest, only about 205.0 g C·m~(-2) ·a~(-1). An obvious seasonal variation of the NPP was observed. The NPP in growth season (from April to October) took about 91.5% of the total in the year, and the peak occurred in mid-July to mid-August, amounting to 37.8% of the total.

Estimation of vegetation net primary productivity on North Shaanxi Loess Plateau

Based on the data from MODIS (Moderate-resolution Imaging Speetroradiometer) and meteorological observatories, and by using improved CASA (Carnegie-Ames-Stanford Approach) model, the vegetation net primary productivity (NPP) and its spatiotemporal distribution on the North Shaanxi Loess Plateau in 2005 were simulated and analyzed. Comparing with the traditional CASA model which only uses a universal mean annual maximum light use efficiency (LUE), the estimated regional NPP by the improved CASA model was more precise, because this improved model used the LUE parameters of different vegetation covers. The detailed land cover classifica-tion also contributed to the increase of the precision via introducing the time-series Normalized Different Vegetation Index (NDVI) and ground survey data to modify and adjust the original clas-sification system based on vegetation map (1: 1000000). The testing of the simulation results from different models with the ground survey data in North Shaanxi showed that the estimation by the modified CASA model was much closer to the real survey data, implying the potential practi-cal significance of this model in estimating the vegetation NPP in North Shaanxi Loess Plateau and its adjacent areas. In 2005, the NPP in North Shaanxi was estimated as 4. 76×1013 g C, ac-counting for about 1.5% of China' s terrestrial total NPP, and the mean NPP was 447.3 g C·m~(-2)·a~(-1), being much higher than that of China' s terrestrial vegetation (323.8 g C·m~(-2)·a~(-1)) in 1982-2000. The spatial distribution pattern of the vegetation NPP showed an apparently declining trend from the southeast to the northwest, with the highest value of 1087 g C·m~(-2)a~(-1) occurred in the broadleaved-and conifer-mixed forests of Huanglong Mountain in southeast part of the region. The mean NPP of desert vegetation in the whole region was the lowest, only about 205.0 g C·m~(-2) ·a~(-1). An obvious seasonal variation of the NPP was observed. The NPP in growth season (from April to October) took about 91.5% of the total in the year, and the peak occurred in mid-July to mid-August, amounting to 37.8% of the total.