Drought Risk Assessment Model And Its Early Warning Mechanism

In recent years, with the growing economy, the global environment are deteriorating, the shortage of water resources are getting more and more serious, drought problem is increasingly prominent and it has attracted widespread meet and becomes scholars urgent need to solve the problem. Qiqihar district was located in the north of SongNen plain, genera of temperate continental monsoon climate, drought happened often, not only in the destruction of ecological environment, but also influenced the region’s agriculture and economic development. Therefore, a systematic study of the drought of Qiqihar region was important. This paper had a more comprehensive introduction to the agricultural drought situation and development trend and introduced modern assessment, prediction, early warning and set up agricultural drought disaster warning mechanism, studied the Qiqihar areas of agricultural drought disaster status and its future development trend.The concrete content included the following:1) Based on reference a vast amount of literature at home and abroad, discussed the basic concept of drought, pointed out that drought should be taken into two aspects,water supply and demand and should pay attention that drought was an anomal phenomenon.Based on the above ideas, this paper improved the definition of the drought and agriculture drought. The refined definition not only considering the natural factors, also the human element. Not only considering the water supply condition, also the water demand. Emphasize the drought was an anomal phenomenon, should descripe the drought as normal standard, its more realistic than description of drought as ideal situation for the standard.2) This paper used Mann-Kendall method and linear tendency estimate method to analyze the change characteristics of average annual rainfall and the annual temperature trend in qiqihar. Used Mann-Kendall method showed that Qiqihar area had the significant temperature rising trend, but the trend of the precipitation was not clear. Used linear tendency analysis method showed that rainfall sequence had long-term decline trend and the temperature sequence had long rising trend. At the same time, using a single factor index, surface precipitation, dry index and Z index respectively established the single index evaluation model to analysis the drought conditions of Qiqihar region from1958to2009.3)This paper set up the drought index system in qiqihar, the drought was classified into five grade levels and established level evaluation model which based on logistic curve of the projection pursuit. In2008, for example, the paper assessed drought level of10areas in qiqihar, the results showed that qiqihar drought was in a serious condition in2008, drought conditions of five areas in ten were severe drought in2008,3areas in ten were great drought, we can see the seriousness of drought.4) This paper set up the regional drought warning index system, summaried and expounded the concrete method of forecasting warning index, determined warning degrees criteria for the classification and warning limit of area drought warning. In qiqihar area as an example, the paper established early warning model based on the BP neural network, the results indicated that,2012was severe drought year, warning signal light was yellow;2015was severe drought year, warning signal light was yellow;2019was partial drought year, warning signal was blue. At the same time, the paper established early warning model based on projection pursuit regression, the results indicated that,2012was severe drought year, warning signal was yellow;2015was partial drought year, light up the blue lights;2019was partial drought year, warning lights was blue. The results indicated that,2012was severe drought, light up the yellow lights;2015was partial drought year, light up the blue lights;2019was partial drought year, warning light was blue.