Analysis And Prediction Of Cultivated Land Pressure On Different Time Scales In Hubei Province Based On EMD

Food is the first priority for the people,arable land is the material basis of food production.With the acceleration of the reform and opening-up process,a large number of cultivated land has been occupied by construction,the contradiction between human and land is extremely prominent,the cultivated land is not of high grade,the distribution is fragmented,and the reserve resources are insufficient,coupled with the influence of natural and man-made activities,resulting in the continuous reduction of high-quality cultivated land resources,the cultivated land resources in China are facing great pressure.The pressure of cultivated land resources is influenced by many factors,which shows typical non-linear characteristics.Empirical Mode Decomposition(EMD)is a better method to deal with non-stationary and non-linear signals,which can overcome the shortcoming of traditional analysis method without time domain resolution,a more appropriate approach.The cultivated land pressure index of Hubei Province from 1978 to 2018 was calculated by using the revised cultivated land pressure index model,and the characteristics of time series and Barycenter migration of cultivated land pressure were analyzed.The empirical mode decomposition(EMD)method was used to decompose the cultivated land pressure index from 1978 to 2018,and the contribution rate of period and variance was obtained by Hilbert transform.Using the method of grey relational grade analysis,11 indexes closely related to the pressure index of cultivated land were selected.After data standardization,the driving factors were decomposed by EMD,and the driving factors were determined according to the contribution rate of period and variance,then,the wave spectrum of different scales was plotted,and the optimal equation of the pressure change of cultivated land was constructed by using the stepwise regression equation.The empirical mode decomposition(EMD)component is added to the dynamic prediction model,and the prediction of the cultivated land pressure index is made from the two angles of the original and the change of the growth rate respectively,and is compared with the traditional grey prediction model,implement precision verification.This study′s conclusions are obtained:(1)The cultivated land pressure index of Hubei Province from 1978 to 2018 was calculated by using the improved cultivated land pressure index model.In terms of time characteristics,the pressure index of cultivated land fluctuated from 1978 to 2003,and the pressure increased gradually.It reached a peak of 1.163 in 2003,and the pressure of cultivated land was low,the actual per capita cultivated land area is larger than the minimum cultivated land,the cultivated land resources basically meet the needs of the population for food production and life in the region,and the cultivated land pressure is no pressure.From 1978 to 2018,the center of gravity of cultivated land pressure moved from Xiaogan to Wuhan,from(114.245°N,31.591°N)to(114.339°N,30.229°N),the accumulative moving distance is 151.66 km,and the speed of Barycenter moving is the fastest from 1996 to 2006,which is 11.91 km/year.(2)The pressure index of cultivated land from 1978 to 2018 was decomposed into two intrinsic mode functions and one trend function by EMD.On the scale of IMF1,there are about 16.7 cycles with an average period of 2.4 years,of which the period is long and the amplitude is large in 1978-1995,then the frequency is faster,the period is shorter and the amplitude is smaller.On the scale of IMF2,there are 5 cycles,the average fluctuation period is 8.2 years,and the fluctuation frequency is slow first,then fast,and the amplitude decreases from large to small.On the trend scale,the pressure of cultivated land gradually increased from 1978 to 2003,and reached the peak around2003,then,the pressure gradually lightened.According to the variance contribution rate of each scale,the increasing trend is the main trend,and the 8.2-year cycle is the auxiliary trend.(3)The driving factors of different time scales are determined according to the contribution rate of period and variance,and the driving mechanism of different time scales is analyzed by using EMD and stepwise regression equation.The results show that the influence degree of short-term scale factors on the pressure index of cultivated land is as follows: per capita cultivated land area > per capita net income of farmers >mechanization level > Fertilizer Application Amount > grain yield per unit area >Multiple Cropping Index > Irrigation level.In the medium scale,the order of the influence degree of each factor on the pressure index of cultivated land from big to small is: grain per unit area yield > Fertilizer Application Quantity > urbanization rate >per capita GDP > construction occupation ratio > agricultural output ratio.The influence degree of long-term scale factors on cultivated land pressure index from big to small is as follows: grain per unit yield > per capita GDP > per capita net income of farmers > mechanization level > per capita cultivated land area > irrigation level >agricultural output value ratio.In the short term,we should pay attention to the improvement of grain production capacity,carry out the strictest cultivated land protection policy,ensure that the quantity of cultivated land is not reduced,optimize the planting structure and improve the multiple cropping index,raise the level of mechanization,etc.Medium-and Long-term policies pay special attention to socioeconomic factors,such as reducing the use of cultivated land for construction,increasing input in agricultural science and technology,increasing food productivity and reducing the use of chemical fertilizers.(4)The IMF components of each driver were added to the prediction model,and the regression equation was used to predict the 2028 pressure index from 2019 to 2028.According to the forecast of the original annual average growth rate,the cultivated land pressure index will increase first and then decrease in the next 10 years,the cultivated land pressure in 2021-2025,the cultivated land pressure in the 2028 is the warning pressure state,and the rest years are the non-pressure state.If the original growth rate changes and two assumptions are made from the socio-economic and food production point of view,if the assumptions are made from the socio-economic point of view,the pressure index of cultivated land will be reduced,however,the 2021,2023 and 2028 are still more than 0.9,and the cultivated land is unsafe.If we assume that from the point of view of food production,the pressure index of cultivated land will be greatly reduced,but 2024,2025,2028 still exist pressure.Compared with the traditional GM(1:1)model,the precision of the average relative error,RMSE and Mae is lower than that of the GM(1:1)model.