| As the world's most widespread and fragile ecosystem,grassland ecosystems have been highly valued by scholars around the world.Half of the territory of Ningxia is grassland,which is one of the important ecosystem types in Ningxia.At the same time,grassland,as an important strategic resource in Ningxia and the basis for local livestock husbandry,has played an important role in local economic and social development.Ningxia is located in the arid and semi-arid transitional zone,and the grassland ecosystem is particularly fragile.Since the 1980s,some scholars have noticed many problems in grassland degradation,desertification,and biodiversity reduction in Ningxia,but there is still lack of macroeconomic data on grassland productivity in Ningxia.Therefore,the use of remote sensing techniques to accurately monitor large-scale and long-term sequences of grassland productivity,and to further understand the development of grassland ecosystems,has important implications for grassland utilization and scientific protection in Ningxia.This paper used the three interpolation methods of Anusplin,inverse distance weights and spline functions to spatialize the meteorological elements,and compared and analyzed the interpolation accuracy of different methods.At the same time,this paper used the meteorological spatial interpolation data to drive the CASA model to estimate the NPP of Ningxia grassland,and introduces the grassland years of actual measurement data and MOD17A3 data to carry out reliability test on the estimation results respectively.This paper used the three interpolation methods of Anusplin,inverse distance weights and spline functions to spatialize the meteorological elements,and compared and analyzed the interpolation accuracy of different methods.At the same time,this paper used the meteorological spatial interpolation data to drive the CASA model to estimate the NPP of Ningxia grassland,and introduces the grassland years of actual measurement data and MOD17A3 data to carry out reliability test on the estimation results respectively.Based on this,The multi-year trend of grassland in Ningxia during 2000-2015 and its response to temperature and precipitation changes were analyzed and the main limiting factors for grassland growth were explored.The main conclusions of the study were as follows:(1)In the three interpolation methods,the "Bull's Eyes Effect" appeared in the inverse distance weight interpolation method.The other two interpolation methods can simulate the spatial gradient distribution characteristics of temperature and precipitation appropriately in Ningxia,and the Anusplin interpolation method obviously had higher interpolation accuracy.Additionally,the interpolation method of Anusplin obviously had higher interpolation precision,among which the interpolation accuracy of temperature was the most high,and interpolated temperature changes were the most delicate.(2)Meteorological spatial data based on Anusplin interpolation method and IDW interpolation method can drive the CASA model,and can estimate the NPP of Ningxia grassland.The NPP value estimated by the Anusplin meteorological interpolation spatial data driven CASA model had the highest correlation with the average NPP measured over many years and the total grass production in the entire region.Its NPP estimation was closer to the actual situation,which further showed that improving the interpolation accuracy of meteorological elements can improve the accuracy of the NPP estimation of the CASA model to some extent.(3)Using the MOD17A3 NPP data as validation data to test the estimation accuracy of different types of model,the results showed that the CASA model driven by Anusplin meteorological interpolation spatial data has high precision in estimating NPP in dry grassland,shrub grassland,dry desert grassland and desert grassland.The error of NPP estimation of marsh grassland and mountain meadow was large and the precision was high,and their accuracy needs to be improved.(4)The NPP results simulated by the CASA model in recent 16 years showed that the average NPP of the grassland in the region for many years is 148.28g·Cm-2·a-1,but the spatial heterogeneity was large,and the spatial distribution pattern showed a clear north-south high middle low characteristic.In addition,the differences in NPP between different types of grassland are also large,among which the meadows in mountain meadows was the grassland types with the highest NPP in Ningxia grassland,and the NPP value was 518.34 g·Cm-2·a-1.(5)The trend analysis showed that the average annual NPP of Ningxia grassland in the past 16 years was in a fluctuating upward trend,with a linear growth rate of 3.84 g·Cm-2·a-1(P<0.01).The grassland area with NPP? 100 g·Cm-2·a-1 fluctuated and decreased,while the grassland area with NPP?300 g·Cm-2·a-1 fluctuated and decreased.Spatially,the average annual NPP increase rate of grassland in Ningxia decreased from south to north,and NPP in 90%of grassland showed an upward trend,of which 61%of grassland showed a significant upward trend(P<0.05),including desert grassland,steppe,and shrub grass.Hurst index analysis found that the NPP trend of most grassland in Ningxia had strong continuity.(6)The correlation analysis between grassland NPP and meteorological factors indicated that the annual NPP of Ningxia grassland was mainly affected by annual precipitation on the pixel scale,and grassland in the south and central regions had the strongest response to annual precipitation.There was no time lag in the correlation of total monthly precipitation and NPP of grassland in growing season,but there was a one-month time lag with monthly mean temperature. |
PDF Full Text Request