| Agrometeorological disaster is the principal natural disaster damaging agricultural production, which is closely linked with agricultural economic benefits. The impact of agrometeorological disasters on the agricultural economy must be considered in adjustment of cropping structure and addition of improved crop varieties, which would reduce that impact. Tea is one of the "Top Ten leading agricultural products" in Zhejiang Province, and tea production has been the main income source for farmers in mountainous areas. The earlier that spring tea begins to sell, the higher its price. For that reason, tea farmers choose early varieties without additional consideration. This causes low temperatures and frost in spring to be the main limiting factors of spring tea production in the province. Frost disaster monitoring and frost risk analysis for tea production could provide a meteorological basis for development of the tea industry based on local conditions, and increase the income of tea producers. The present work uses temperature data from automatic meteorological stations collected since2004. Based on relationships between tea production and meteorological conditions, we established a model for assessing the economic loss rate of tea production caused by frost disaster. This economic loss rate was evaluated using both the model and remote sensing technology. Finally, using the model and meteorological data, we analyzed the frost risk of tea production. Results were as follows:1. Using meteorological observation data from the county meteorological station as input to a support vector machine, along with daily mean and daily minimum temperature differences of that station and automatic meteorological stations as outputs, regression models of temperature data from automatic meteorological stations were constructed through the method of s-SVR with RBF kernel function. Model-fitted values agreed well with actual ones, and mean absolute error was less than0.5℃. Using meteorological data of the Shenzhou, Shaoxing, and Shangyu meteorological stations for testing, results showed that the support vector machine had good stability and precision in correction of daily mean temperature and daily minimum temperature from January through April. Absolute errors of fitted extreme minimum temperature were mostly less than1℃during low-temperature and frost disaster periods, which satisfies the requirements of frost disaster evaluation and risk regionalization. Based on this finding, daily mean temperature and daily minimum temperature series from automatic meteorological stations during January through April could be corrected to the prior30-year period.2. Based on linear regression, forecast models of beginning dates of tea picking (BDTP) for teas "Wuniuzao,""Longjing43," and "Jiukeng" were constructed. Among those, the BDTP of Wuniuzao is between the last ten days of February and middle ten days of March, when temperature varies greatly in Xinchang County. Furthermore, because continuously cold and rainy weather in some years can delay BDTP, accumulated temperature should be considered in Wuniuzao BDTP forecasting. Differences of BDTP forecasting model-fitted and actual values were±1d. Using tea phenological phase, production and tea price data, models of tea buds and leaves growth, picking amount of tea buds and leaves and tea prices were established. Based on models of picking amount of tea buds and leaves and tea prices, fitting models of economic output of the three tea species were established, and economic output fitted with meteorological data in each day. Based on the frostbite rate of tea buds, we defined six grades of tea frost disaster. Minimum temperature was the main meteorological impact on this grade. On sunny and windless mornings, minimum temperature of the tea canopy was less than minimum air temperature, by2℃to4℃. A tea weather frost index was established by combining frost observation and investigation data. During frost disasters, fitted values of tea-bud frostbite rate, based on the frost index and minimum air temperature, were similar to actual values.Based on the BDTP, tea economic output rate and tea frost weather index, we established an evaluation model of tea economic loss rate caused by frost. Frost evaluation from9to11March2010gave results similar to those of the investigation by tea farmers at the meteorological stations; these results can be used in the tea meteorological service business. However, tea plants grow in mountainous regions, at elevations higher than those of the automatic meteorological stations. Therefore, minimum air temperature in the tea production area is colder than at those stations. Consequently, the economic loss rate of tea production in the entire county is slightly greater than that from the evaluation.3. Tea frost disasters occurred on sunny and windless mornings, when tea buds suffered frostbite because of minimum air temperatures below0℃, caused by surface radiational cooling. Therefore, remote sensing technology may be used to monitor tea frostbite. First, based on relationships of BDTP with geographic factors, BDTPs were calculated on grids. Next, surface temperature was retrieved with four split-window algorithms and National Oceanic and Atmospheric Administration/Advanced Very High Resolution Radiometer (NOAA/AVHRR) data. Based on relationships between minimum air temperature, retrieval surface temperature and geographic factors, minimum air temperatures were retrieved on all grids. Using the retrieved minimum air temperature, BDTP and the assessment model of tea economic loss rate caused by frost disaster, we evaluated tea economic loss rates caused by such disaster on all grids. The evaluation from9to11March2010showed that the evaluated value was similar to the measured value.4. Using the probability of frost and tea economic loss rate caused by frost disasters in spring tea production as the analysis object, tea-tree frost risk was determined. Based on the relationship of BDTP probability of each day with geographic factors, the relationship of probability of low temperatures less than or equal≤0℃on each day with geographic factors, combined geographic factors and evaluation model of tea economic loss rate caused by frost, tea-tree frost risks were calculated on all grids. Frost risk of tea Wuniuzao was40-100%, Longjing43was0.8-7.7%, and Jiukeng was less than1.26%. Based on frost risk of Wuniuzao tea tea, Xinchang county was divided into four areas for tea frost risk.Based on the frost risk analysis, a scheme for tea frost weather index insurance was developed for each town. Frost risk of Wuniuzao was very high, so its deductible was high. |
PDF Full Text Request