内蒙古锡林郭勒盟草原产草量动态遥感估算

Remote Sensing Dynamic Estimation of Grass Production in Xilinguole,Inner Mongolia

草原产草量的监测是草地资源空间动态研究的重要衡量指标,是草地资源合理利用和载畜平衡监测的重要依据.基于371个样地调查数据和2005～2009年的MODIS-NDVI遥感数据,建立地面样方的产草量与遥感数据的关系模型,模拟分析了内蒙古锡林郭勒盟草原产草量的时空分布.结果表明:(1)建立的各模型方程均有较好的相关关系,其中幂函数的相关关系最优,通过预留样方数据的验证,模型精度为78%,幂函数模型作为遥感估测应用可行;(2)锡林郭勒盟草原的产草量5年平均为3455万吨,折合干草总量为1112万吨,平均单产为567.23kg/hm2,草原产草量的空间分布呈东高西低的格局;(3)2005～2009年,锡林郭勒盟草原产草量有明显的波动,干草变化范围为800～1400万吨,变异系数为20.42%;(4)不同草地类型的产草量及其年际间变化存在较大的差异,荒漠类草原产草量低,年际间变化较大;草甸类草原产草量高,年际间变化相对较小.草原产草量的时空变化还与降水量、气温等主要气候因素关系密切,特别是受降水量的时空变化影响显著.研究结果可以为中国草地资源的保护及合理利用提供参考依据.

The monitoring of grass production is not only an important measurement index for researching the spatial dynamic patterns of grassland resources, but also an important basis for the optimal use of grassland resources and for monitoring the balance between grassland forage supply and livestock demand. A total of 371 grassland investigation and remote sensing （RS） data of MODIS-NDVI from 2005-2009 were used to establish a relational model between grass production and RS data. The temporal and spatial distributions of grass production in Xilinguole, Inner Mongolia were simulated and analyzed. The main conclusions are as follows： （1） Each model equation had good relativity, among which the power function model was the most appropriate for the calculation of grass production. Through reservation quadrat data, the accuracy of the production estimation model was 78%. It was feasible for the power function model to be applied for RS monitoring. （2） Grass production in the Xilinguole grassland amounted to an average of 34550000 tons over five years, equivalent to 11120000 tons of hay, and the dry weight per unit area was 567.23 kg/hm2. The spatial distribution of grass production showed a ＂high east and low west＂ pattern. （3） An obvious fluctuation in grass production appeared during 2005-2009. The range of hay was 8000000-14000000 tons, and the coefficient of variation was 20.42%. （4） There were greater differences in the single grass yield and the annual variation of different grass types. There were lower grass yields and greater annual variation in desert grassland, and higher grass yields and relatively lesser annual variation in meadow grassland. The spatial and temporal variation of grass yields also had the close relation with the major climate factors （precipitation, temperature, etc.）, especially with the influence of precipitation. This research provides a reference for the protection and utilization of grassland resources.