Study on development of agricultural system model on a large scale

大规模监测作物状况和产量估计对粮食供应很重要。研究目标是：(1)利用地球静止气象卫星太阳辐射数据和气象格网日、时常规观测数据进行气象值制图；(2)利用作物生长模型与气象数据估算模型耦合模拟作物生长；(3)利用卫星数据在国外应用气象变量估算技术；(4)利用密集观测点监测农业生产环境信息。主要结果如下：(1)利用晴空识别算法和土地利用数据，改进了现有的太阳辐射估算模型，均方根误差为16%，可以较好地从常规观测数据中估算网格气象数据（气温、风速和日照时数）。(2)我们演示了将作物模拟模型与GMS图像估算的太阳辐射数据和气象数据结合使用。该系统预测水稻产量的均方根误差为65kg/10a。我们期望通过本研究开发的系统，可以为农民提供咨询，分析气象灾害对作物的危害，预测气候变化对作物产量的影响。(3)利用气象卫星资料估算黄河流域的净辐射和净降水，研究了黄河流域因缺水而导致的农业减产。所得结果令人满意。(4)利用气象观测机器人网络数据，以250m栅格获取小麦成熟期预测图，建立基于霜冻概率的作物霜冻危害风险图。

Monitoring crop condition and production estimates on a large scale is important for food supply. The objectives of the study are (1)mapping meteorological values by use of Geostationary Meteorological Satellite data for solar radiation and by routine observation data for meteorological grid data on daily and hourly basis, (2)simulating crop growth by coupling of crop growth model and estimation model for meteorological data, (3)applying estimation technique for meteorological variables by satellite data to a foreign country, and (4)monitoring environmental information for agricultural production by dense observation points.The main results are as follows. (1)We have improved existing model for estimating solar radiation by use of clear sky recognition algorithm and land use data with RMS Error of 16%, and could estimate grid meteorological data (air temperature, wind speed and sunshine duration) from routinely observed data fairly well. (2)We have demonstrated the use of a crop simulation model together with solar radiation data estimated by GMS images and meteorological data. The system showed rice yield predictions at RMS Error of 65kg/10a. We expect that the distribution of crop growth and crop yield by the system developed in this study will be used for consulting farmers, analyzing crop damage by meteorological disaster and predicting crop yield change by climate change. (3)We have applied the method for estimating net radiation and precipitation by meteorological satellite data on Yellow River basin in China where water shortage is reducing agricultural production. The obtained results were satisfactory. (4)By using the data of a meteorological observation robot network, the wheat maturity period prediction map could be obtained by the grid of 250m, and the risk map of the frost damage of crops based on frost probability has been created.