Abstract: 【Objective】Satellites and ground multi-source data were used in this study to achieve dynamic monitoring of cold damage of double cropping rice by combining with extraction result of double cropping rice planting area, dynamic development phase maps, daily mean temperature maps and monitoring index of cold damage. The result can provide technical supports for the agricultural production sectors to make quick response to cold damage and to reduce the loss of cold damage.【Method】Using daily mean temperature of 190 meteorological stations located in southern China from 1951 to 2011, the annual average accumulated temperature (≥10℃) was obtained for each station. According to correlativity between annual average accumulated temperature and geographical factors, the accumulated temperature prediction model was established, and according to the model, the spatial distribution map of accumulated temperature was obtained. Using accumulated temperature threshold of 5 300 ℃.d for double cropping rice production, the area that average accumulated temperature less than the value was removed from research area. The result was as accumulated temperature zone of double cropping rice. The 10 years distribution of grain crops was obtained by using the classification products of MODIS MCD12Q1 from 2001 to 2010. Finally, the information of double cropping rice planting area was extracted by calculating the intersection of accumulated temperature zone of double cropping rice and grains classification products of MODIS MCD12Q1. This research used development phase data of 167 agricultural experiment stations in southern China from 1981 to 2011, according to the correlation between many years average date of development phase and geographical factors, the regression fitting distribution map of development phase was obtained for double cropping early rice. On the basis of residual error correction of the fitting results, the maps of dynamic development phase were made by using two static maps before and after through EVNI + IDL programming. Multivariate regression model of daily mean temperature was established by using the correlation between AMSR_E_L2A 89 GHz bright temperature, geological factors and daily mean temperature of meteorological stations, as a result to complete the calculation of February 2010 to May daily mean temperature. 【Result】Using 90 meteorological stations observation data that were not involved in modeling to validate the accumulated temperature prediction model, and the results showed that there was no significant difference between estimated and measured values, the root mean square error is 289℃. The estimate accuracy can meet the needs of the research. By calculating the intersection between the accumulated temperature zone of double cropping rice and MODIS grains classification products, the extractive planting area of double cropping rice accounted for 29.26% of total of the study area. Through the steps, on the one hand, improved the use value of monitoring results of cold damage. On the other hand, can make the real-time monitoring of cold damage of double cropping rice be launched before haven't get the accurate planting location information about double cropping rice; Using observation data of 59 agricultural experiment stations that were not involved in modeling to validate development phase simulated results of double cropping early rice seeding and reviving period, the results indicated that the root mean square errors were, respectively, 5.68 days and 4.24 days between real and fitted many year average date of development phase. Model accuracy meets the requirements, the experimental results can be used to subsequent analysis; Daily mean temperature was obtained by hybrid modeling method, the method makes full use of the characteristics of microwave remote sensing, the data is affected little by the weather, and can overcome the south year-round cloudy weather conditions, root mean square error (RMSE) of simulative daily mean temperature is 1.69℃. Estimated precision of daily mean temperature was increased by 0.35℃ because of the addition of microwave data. Monitoring results for 2010 spring cold damage of double corpping rice highly corresponds to reported content of agricultural statistic department in time and space.【Conclusion】The method constructed by this paper can realize the large-scale synchronous monitoring of cold damage of double cropping rice. The method has universality. It can also be applied to large scale cold damage dynamic monitoring of other crops.

Key words: double cropping rice, spring cold damage, dynamic monitoring