Research On Intelligent Method Of Identification,evaluation And Early Warning Of Regional Agricultural Drought Risk

Drought disaster has long been a major type of natural disasters affecting the economic and social development of mankind.The territory of China overlaps with two well-known natural disaster belts on the earth.One is on the Pacific Rim and another one is the 20° to 50°N belt.Constrained by its geographical environment and climate characteristics,drought events occur frequently.With the aggravation of human activities,the drought situation is becoming more and more serious;it has brought huge losses to the production and human life.The problem of drought poses the most direct and serious threat to the security of agricultural production in China,which makes the research on regional agricultural drought risk system become an important issue to realize the national drought reduction strategy and a positive measure to improve the comprehensive disaster forecasting and early warning ability from the perspective of scientific initiative.In view of this,this paper takes the Huaibei Plain of Anhui Province as the research area.Based on the theory of natural disaster risk system,the constituents of regional agricultural drought risk system were identified from four aspects i.e.disaster sensitivity,disaster factors and exposure,and drought resistance.Based on the origin and the process of the disaster loss,this paper carried out an empirical study to evaluate the risk curve evaluation method combined with field experiment and crop growth model.The intelligent methods such as set pair uncertainty analysis,genetic optimization analysis,fuzzy set analysis and grey correlation analysis were used to establish the quantitative evaluation and early warning models.The main conclusions are as follows:(1)The paper identified the constituents of regional agricultural drought risk system and analyzed the key parameters in the system.It is suggested that the short duration meteorological drought events occur frequently in the investigated area,yet the drought intensity is not severe.The spring drought and the spring-summer continuous drought were easy to occur.In recent years,drought disaster is higher than other natural disasters,and drought is still the main one.Huainan and Fuyang were high disaster sensitive areas in the region.Bengbu,Fuyang and Bozhou have a weak agricultural drought resistance,which needs to be improved.The results provide data basis for the follow-up research of agricultural drought risk assessment in the study area.(2)The risk curve evaluation method of regional agricultural drought based on the origin process of disaster loss has been developed.Combined with the results of risk identification,the meteorological drought events which coincide with the growth period of summer maize were selected for joint distribution analysis of drought characteristic variables.The parameters of the Aqua Crop model were adjusted and verified by using the observation data of summer maize field test in 2018 and 2019.The quantitative relationship between drought frequency and yield reduction rate of maize under different irrigation levels was obtained.The rationality and effectiveness of the regional agricultural drought risk curve assessment method has been proved by the results.A scientific basis for the effectuation of disaster prevention and mitigation measures in Bengbu is provided correspondingly.(3)A regional agricultural drought risk assessment model was established based on set pair uncertainty analysis and numerical simulation of disaster loss risk.It is considered that the recurrence period of regional agricultural drought in the research area was about 1-3 years.With the increasing drought intensity,the frequency of the drought occurrence decreased significantly.The recurrence period of severe drought and above was about 10-30 years.The frequency of severe drought and above in the eastern region was higher than that in the western region.The risk for different regions from high to low was Huainan,Bengbu,Suzhou,Fuyang,Bozhou and Huaibei.The results can provide effective reference in similar areas and promote the quantitative study of drought risk.(4)A regional agricultural drought risk early warning model based on early warning system theory and comprehensive early warning index was established.The early warning model is composed of single early warning effect evaluation based on the improved KLR model,early warning signal design based on comprehensive early warning index and prediction based on GM(1,1).The corresponding warning signals were given to each city in the study area.The results show that in the case of short-term forecast of the annual growth rate of warning indicators,the early warning levels were basically from no alarm(green)to light alarm(blue),and the comprehensive early warning index values of Huainan are relatively large,which should be paid reasonable attention to.The results will help to strengthen the promotion and application of early warning methods from the practical level,and act as valid methods to improve the disaster response ability of regional agricultural production.