| Agricultural non-point source pollution(ANSP)is one of the important causes of water environmental pollution in various countries.In the process of water pollution control,the long-term main concern is the source pollution,and the agricultural non-point source pollution is ignored.According to the results of the second pollution source survey released in China in 2020,the contribution of ANSP to water pollution is very high.At present,more and more attention has been paid to ANSP.The Chinese government has issued a number of documents on the prevention and control of ANSP.And it is proposed that due to the unique geographical environment and economic characteristics of the Yangtze River Economic Belt,ANSP is heavy,and it is necessary to accelerate the regional governance of ANSP.As a connecting point one of the"Belt and Road"and the Yangtze River Economic Belt,Chongqing has an important ecological status in the national regional development pattern and Upper Yangtze River.Chongqing has the background characteristics of high proportion of rural areas,high intensity of fertilizer and pesticide application,high proportion of hilly and mountainous areas,large and concentrated rainfall,which leads to large potential,strong driving force and wide range of ANSP.At the same time,combined with the Chinese government's attention to the application of grid management in modern agricultural management,grid management will be an important means to solve the risk control and information construction of ANSP at multi-level scale.Therefore,in order to ensure the ecological security of the region and the Yangtze River Basin,Chongqing needs to carry out the construction of ANSP risk measurement and visualization platform under grid management mode to improve the risk prevention,control and information ability of ANSP.By combing the research hotspots,trends and problems of ANSP risk measurement,we find that there are some key problems,such as the disunity of research path of ANSP risk measurement at multi-level scale,the inconsistency between the visualization ability of measurement results and the needs of managers.Therefore,taking Chongqing's multi-level scale range as the research object,we focus on the key issues to carry out the research of ANSP risk measurement and visualization under multi-level grid.The main contents are as follows.(1)The county grid scale is the macro research scale,which focuses on identifying the spatial-temporal evolution trend of ANSP risk in different regions in a large scale.Taking Chongqing as the research object and supported by GIS technology,we build a risk measurement PTA3D model covering three dimensions of pressure kinetic energy,transformation kinetic energy and absorption kinetic energy.And reveal the spatial-temporal evolution law of ANSP by using the methods of barycenter and kernel density analysis,and discuss the accuracy and influencing factors of risk assessment.(2)The township grid scale is the macro policy landing level,which focuses on identifying the evolution trend of ANSP risk under the influence of future development mode selection.Taking Fuling District as the research object,we simulate the future land use change under the scenario of natural development and cultivated land protection by combining CLUE-S model,and use the output risk model to reveal the spatiotemporal evolution law of risk probability of each grid,and explore the response of risk degree to the change of output coefficient.(3)The village grid scale is the specific implementation level of policy,with the emphasis on identifying risk resistance,transport path and landscape optimization effect of ANSP.Taking Nantuo area as the research object,based on the daily water level data,we carried out the interpretation of spatial factors such as water level line and land use type in the submergence and non-submergence period.And the minimum cumulative resistance model is constructed to identify the resistance surface,risk level and transmission path of ANSP in different periods by introducing the source sink process,and discusses the role of resistance model and landscape optimization in risk prevention and control.(4)The block grid scale is the operation level of farmers,which focuses on the analysis of the specific practice and optimal control benefits of field management of farmers'behavior supported by high-precision data.Taking Muhe village as the research object,we introduced low-flying remote sensing(RS)UAV multispectral technology to obtain six kinds of high-resolution RS data such as image and NDVI index.And introduced random forest algorithm to carry out feature classification and block division,determined fertilizer and pesticide application situation of each feature through farmer behavior investigation.And discussed the accuracy of random forest algorithm,the influencing factors of ANSP risk and the effect of reducing engineering.(5)Based on the risk assessment data of ANSP in multi-level grid scale,we construct the risk visualization platform of ANSP under multi-level grid based on the differences of management requirements in each scale under the framework of B/S,and discuss the risk management and prevention and control strategies.The main conclusions of this paper are as follows.(1)The ANSP risk measurement PTA3D model constructed at county scale could reflect the real situation,and the geographic methods such as barycenter and kernel density analysis could effectively identify the risk evolution.The results showed that:(1)ANSP risk in Chongqing has an obvious trend from low risk to high risk(the proportion of high and extremely high risk increased from 17.82 and 16.63 in 2000 to 18.10 and 16.76 in 2015,respectively).(2)The risk of metropolitan region was higher than that of southeast and northeast in Chongqing.The distribution of the barycenter in the risk areas from high to low levels was characterized by the distribution from west to east.The high and extreme high risk areas showed obvious eastward transfer trend(from2000 to 2015,the barycenter of the high and extreme high risk areas moved 4.63 km and 4.48 km according to 1.68°and 12.08°east by north,respectively).(3)The results of kernel density showed that the high risk was mainly concentrated in the metropolitan area,the risk in the northeast and southeast of Chongqing has a low degree of agglomeration and an upward trend,and the spatial fragmentation of the agglomeration area of Chongqing was increasing and the trend of decentralization and concentration appears.(4)The water quality of the river and the area composition of land types within a certain range of the shoreline were analyzed with the model measurement results,and it was found that the model measurement results could reflect the real ANSP risk situation of the region.(5)The analysis of risk driving showed that the composition of land types has a significant impact on the risk intensity,and the risk of regions with high proportion of"source areas"was significantly enhanced.The occupation of cultivated land in urban areas in the process of urbanization was one of the factors to promote the eastward movement of the barycenter of high and extremely high risk regions.(6)Through this research path,we identified the high and extremely high risk concentration areas as key areas for prevention and control,which could effectively reduce the risk of ANSP.(2)The CLUE-S model introduced from the township scale has achieved good results in the future land use scenario simulation,and provides a good support for the risk measurement,risk evolution and influencing factor analysis of ANSP.The results showed that:(1)The risk probability of each town and sub-basin in Fuling District decreased significantly during 2010-2015,especially near the Damu and Wuling mountains.(2)The kappa coefficient of land-use scenario simulation combined with CLUE-S and Markov model was 0.75,which was highly consistent.(3)The future scenario simulation was helpful to identify the TN and TP output risk probability evolution of towns and sub-basins in different development scenarios,and the response relationship between risk probability and land-use.The risk of villages and towns in the north of Yangtze River and the west of Wujiang River was obviously high.For example,sub-basins No.148 and 150 have an increasing risk trend under both development modes.(4)The output risk level of sub-basin has a good response to the change of cultivated land output coefficient(i.e.the adjustment of fertilizer application level).For example,sub-basins No.3,75,104,141,202,211,259,292,330,398 and461 responded to the reduction of output coefficient.(6)The risk characteristics of towns and sub-basins identified by this research path could be effectively applied to the future land structure optimization or fertilizer and pesticide application level control,so as to achieve the purpose of ANSP risk prevention and control.(3)The minimum cumulative resistance model introduced in village scale was helpful to identify the risk resistance value,risk area,transmission path and analyze the landscape optimization effect of ANSP.The results showed that:(1)The minimum cumulative resistance model based on the"source-sink"theory has a good effect in identifying the spatial location of different risk levels,transportation path and key areas of risk prevention and control.(2)The analysis of resistance surface showed that the resistance value was"low in the west,high in the east,scattered in the high value and continuous in the low value".The highest resistance value was a2(>25°cultivated land)>a6(?2°cultivated land)>a3(15-25°cultivated land)>a5(2-6°cultivated land)>a1(rural residential area)>a4(6-15°cultivated land).(3)Under the condition of seasonal water level fluctuation,compared with the non-submerge period,the risk of ANSP was greater,the resistance of transmission path was smaller,and there were more low-level transmission paths in the submerge period.The number of paths with resistance value of level 1 was a4>a5>a1>a3>a6>a2.(4)The transformation from"source land"to"sink land"in different areas around the water body plays an important role in risk control.Under the scenario of farmland adjusted to woodland within 50m and 100m around the water body,the number of a2 transport paths in level 1 was reduced by 13.79%and 53.66%respectively compared with the current situation.(5)The risk distribution and transportation path identified by this research path could be combined with landscape optimization to effectively prevent and control ANSP.(4)The low-flying RS UAV technology,random forest algorithm,farmer survey and other methods introduced in the block scale have achieved good results in the classification of surface features,load risk measurement,and benefit analysis of prevention and control measures.The results showed that:(1)The centimeter level high-resolution data obtained by low-flying RS greatly improves the precision of basic data and the purity of information in grid.The precision of fine ground feature classification under random forest algorithm was as high as 90.05%,which was conducive to the block grid division.(2)Based on the survey of farmers'behavior,we founded that the decline of labor force,the aging of population and the high fertilizer and pesticide application of crops lead to the high risk of ANSP.(3)It was estimated that the proportion of blocks with TN and TP application intensity lower than 200 kg/hm~2 was low,and the monitoring results of water quality also reflect that the regional ANSP was relatively serious.(4)We analyzed the ANSP risk reduction capacity of constructed wetland,and found that it could significantly reduce pollutants and produce better economic benefits.The suspended matter,total phosphorus,total nitrogen,ammonium nitrogen,nitrate nitrogen and COD in the area were reduced by 86.67%,54.66%,81.11%,10.67%,83.85%and 59.42%respectively after being treated by constructed wetland.(5)Through this research path,we could identify the behavior of farmers,block risk characteristics and the reduction capacity of constructed wetland,which could be well applied to regional ANSP risk prevention and control.(5)We integrated the results of ANSP risk measurement and management requirements in multi-level grid scale,carried out the construction of visualization platform,and achieved good results.The results showed that:(1)The establishment of an information-based and visual risk prevention and control platform of ANSP could effectively improve management efficiency and prevention and control capability.(2)The development of visual platform based on B/S framework and Web Storm tool has the advantages of rich and diverse technology,light and fast process,and beautiful visual effect.(3)The visualization platform could well display the ANSP risk measurement results under the multi-level grid,and give better play to the advantages of Geography in the field of Agronomy.The platform could fully reflect the value of geospatial data,and provide digital,efficient and scientific means for management and decision-making.In general,we integrate the disciplines of Agronomy,Geography,Computer Science and others,and consider the different needs of each scale in ANSP risk management,prevention and control,which provides a unified ANSP risk research paradigm for ANSP research path and method under the multi grid scale.And it can effectively solve the problems of disunity of risk measurement path,poor reference of research content and findings,and weak visibility of research results. |
PDF Full Text Request