Effects Of Climate And Terrain Factors On ANPP In Hunshandake Sandy Land

Climate change has become one of the most important environmental problems in the world.It has a certain impact on the ecological environment and social economy.Climate factors such as temperature,precipitation and wind speed are different effects on vegetation productivity in different terrain.Taking Duolun County as the research object,this thesis uses deep learning method to simulate the influence of terrain factors on the spatial distribution of eco-environmental factors(water and heat),and analyzes the influence of different climate change scenarios on net primary productivity(ANPP)through mathematical model.The detailed research methods were as follows:I)through Mann Kendall statistical test and wavelet analysis,the change trend and periodic characteristics of temperature,precipitation and drought degree(based on standardized precipitation index(SPI))in the growing season of 1980-2019 in the study area were analyzed.II)the spatial variation trend and stability of land surface temperature(LST)and soil moisture(Temperature vegetation dryness index(TVDI))obtained from MODIS remote sensing data in the study area from 2001-2019 were analyzed by regression analysis,trend analysis and stability analysis,The spatial variation of LST and TVDI in the region was analyzed by using the natural discontinuity classification method.III)the relationship between the spatial distribution of LST and TVDI and the air temperature and soil moisture of meteorological stations,as well as the topographic factors in the study area were simulated by deep learning method.IV)based on the above deep learning model,the changes of ANPP in the growing season under different climate(temperature and precipitation)change scenarios were analyzed by CASA model.The main conclusions are as follows(1)Climate change characteristics of temperature,precipitation and drought in the growing season of the study area.The temperature increased significantly at the rate of0.042?/a(r=0.685),and the linear trend rate from 1980 to 2000 is 2 times as high as that from 2001-2019,which indicated that the temperature increase in this region had a trend of slowing down.The precipitation shows a downward trend(-0.2884mm/a)from1980-2019.However,this trend is fluctuating.The precipitation shows an upward trend(4.783 mm/a)from 1980-2000,a sharp decline in 2000 and 2001,and then an upward trend(2.284 mm/a)from 2001 to 2019.Accordingly,the trend of aridity in the study area is as a whole(the tendency rate of SPI index is-0.0038/a).From 1980 to 2000,drought decreased(SPI index tendency rate 0.074/a),and extremely severe drought occurred in2001.After that,the humidity showed an increasing trend(SPI index tendency rate0.037/a).Temperature,precipitation and SPI do not show a simple linear change with years,but have some differences in different periods,especially in precipitation and SPI.Time scale should be considered in the analysis of temperature,precipitation and drought.Different time scales is produce different results.(2)The spatial variation characteristics of LST and TVDI.Based on the MODIS remote sensing satellite data from 2001-2019,LST accounts for 66.66%of the total area when 29.05??LST<30.60?.From 2001-2019,the areas of low temperature(23.05??LST<27.65?)and sub low temperature(27.65??LST<29.05?)areas tend to expand;The areas of medium temperature(29.05??LST<29.89?),sub high temperature(29.89??LST<30.60?)and high temperature(30.60??LST<31.24?)had a decreasing trend.TVDI index has a good inversion effect on soil moisture.The soil moisture in the study area had mainly in the normal state(accounting for 50.48%of the total area).The area of normal and humid soil moisture is a tendency to expand,while the area of arid soil moisture is a tendency to decrease.(3)The coefficient of determination(R~2)of simulated and measured LST is 0.8125,the mean absolute error(MAE)is 0.53?,and the mean square error(MSE)is 0.46?.The results show that the main factors affecting the spatial distribution of LST are temperature,NDVI,elevation,solar radiation,surface reflectance,slope and aspect,and their characteristic importance degrees are 0.58,0.238,0.134,0.014,0.013,0.01 and0.008,respectively.When other factors are constant,LST in growing season is 1.7 times of air temperature.With the increase of NDVI,LST decreased,which indicates that vegetation coverage has a certain regulating effect on land surface temperature.LST increases first and then decreases with the increase of elevation and slope,and the change range is small;LST is higher in the East,Southeast and south slope,and lower in the north,northwest and west slope.With the increase of temperature,LST of elevation,slope and aspect increases.(4)When deep learning method simulates TVDI,MAE is 0.06,MSE is 0.01 and R2is 0.7765.The relationship between LST and TVDI was analyzed by quantitative analysis of altitude,NDVI,surface reflectance and other factors.The results showed that:in the growing season,with the increase of LST,soil moisture decreased,which may lead to regional water shortage or drought.TVDI decreased with the increase of precipitation,the greater the precipitation,the higher the soil water content.TVDI first increased and then decreased with the increase of altitude,that is,soil moisture first decreased and then increased.(5)When deep learning method simulates GDD,MAE is 23.8?,MSE is 883.22? and R~2is 0.9217.GDD increases with the increase of temperature,and increases first and then decreases with the increase of altitude and slope;GDD is higher in the East,Southeast and south slopes,and lower in the north,northwest and West slopes.With the increase of temperature,the GDD of altitude,slope and aspect also increased.(6)It is found that the increase of temperature and precipitation has a positive effect on ANPP.Under different hydrothermal scenarios,ANPP increased with the increase of altitude,and decreased first and then increased with the increase of slope;ANPP is higher in the north and northwest slope,and lower in the West and southeast slope.In this thesis,we systematically analyzed the characteristics of water and heat change in time scale and space scale in the study area,and analyzed the impact of climate change on vegetation productivity in different terrains.It overcomes the limitation that remote sensing method can only monitor the current eco-environmental factors,and makes it possible to study the response of eco-environmental factors to climate change in different terrains.The research results can provide theoretical reference for ecological environment protection in the study area under climate change,and have important significance for regional sustainable development.