利用天然草地牧草光谱观测资料、牧草产量资料、气象资料和NOAA／AVHRR资料，建立了天然草地牧草产量光谱植被指数和卫星遥感监测模型、气监测模型，提供及时准确地掌握牧草产量变化的科学手段。建立了天然草地牧草产量遥感预测模型及气象预测模型，可以根据需要提供不同时效的卫星遥感预测结果和气象模型预测结果。气象模型精度较高，但气象站点有限，往往以点代面；遥感技术宏观性强，空间信息丰富，可以弥补气象模型的不足；两者既可以互相验证，又可以取长补短。1995年以后服务表明，这些模型达到牧业气象业务服务的要求。

Using grass spectral data, yield data, meteorological data and NOAA/AVHRR data of natural grassland, grass yield monitoring models using remote sensing and meteorological data are established. These models can provide grass yield during growth period and this is very useful to direct animal husbandry production. The grass yield predicting models using NOAA/AVHRR and meteorological data are established. These models can provide grass yield in advance, and this is very important to arrange animal husbandry production. Meteorological models are more accurate, but these models can only estimate the grass yield at limit locations. Remote sensing models can estimate the grass yield an any location. Therefore, the monitoring and predicting results are reliable if meteorological models and remote sensing models are used together. The estimated results after 1995 show that these models can meet the needs of operational services.