Development of a surface modeling method for mapping soil properties

High accuracy surface modeling （HASM） is a method which can be applied to soil property interpolation. In this paper, we present a method of HASM combined geographic information for soil property interpolation （HASM-SP） to improve the accuracy. Based on soil types, land use types and parent rocks, HASM-SP was applied to interpolate soil available P, Li, pH, alkali-hydrolyzable N, total K and Cr in a typical red soil hilly region. To evaluate the performance of HASM-SP, we compared its performance with that of ordinary kriging （OK）, ordinary kriging combined geographic information （OK-Geo） and stratified kriging （SK）. The results showed that the methods combined with geographic information including HASM-SP and OK-Geo obtained a lower estimation bias. HASM-SP also showed less MAEs and RMSEs when it was compared with the other three methods （OK-Geo, OK and SK）. Much more details were presented in the HASM-SP maps for soil properties due to the combination of different types of geographic information which gave abrupt boundary for the spatial varia- tion of soil properties. Therefore, HASM-SP can not only reduce prediction errors but also can be accordant with the distribution of geographic information, which make the spatial simula- tion of soil property more reasonable. HASM-SP has not only enriched the theory of high accuracy surface modeling of soil property, but also provided a scientific method for the ap- plication in resource management and environment planning.

HASM quantum machine learning

The miniaturization of transistors led to advances in computers mainly to speed up their computation.Such miniaturization has approached its fundamental limits.However,many practices require better computational resources than the capabilities of existing computers.Fortunately,the development of quantum computing brings light to solve this problem.We briefly review the history of quantum computing and highlight some of its advanced achievements.Based on current studies,the Quantum Computing Advantage(QCA)seems indisputable.The challenge is how to actualize the practical quantum advantage(PQA).It is clear that machine learning can help with this task.The method used for high accuracy surface modelling(HASM)incorporates reinforced machine learning.It can be transformed into a large sparse linear system and combined with the Harrow-Hassidim-Lloyd(HHL)quantum algorithm to support quantum machine learning.HASM has been successfully used with classical computers to conduct spatial interpolation,upscaling,downscaling,data fusion and model-data assimilation of ecoenvironmental surfaces.Furthermore,a training experiment on a supercomputer indicates that our HASM-HHL quantum computing approach has a similar accuracy to classical HASM and can realize exponential acceleration over the classical algorithms.A universal platform for hybrid classical-quantum computing would be an obvious next step along with further work to improve the approach because of the many known limitations of the HHL algorithm.In addition,HASM quantum machine learning might be improved by:(1)considerably reducing the number of gates required for operating HASM-HHL;(2)evaluating cost and benchmark problems of quantum machine learning;(3)comparing the performance of the quantum and classical algorithms to clarify their advantages and disadvantages in terms of accuracy and computational speed;and(4)the algorithms would be added to a cloud platform to support applications and gather active feedback from users of the algorithms.

Progress on spatial prediction methods for soil particle-size fractions

Soil particle-size fractions(PSFs),including three components of sand,silt,and clay,are very improtant for the simulation of land-surface process and the evaluation of ecosystem services.Accurate spatial prediction of soil PSFs can help better understand the simulation processes of these models.Because soil PSFs are compositional data,there are some special demands such as the constant sum(1 or 100%) in the interpolation process.In addition,the performance of spatial prediction methods can mostly affect the accuracy of the spatial distributions.Here,we proposed a framework for the spatial prediction of soil PSFs.It included log-ratio transformation methods of soil PSFs(additive log-ratio,centered log-ratio,symmetry log-ratio,and isometric log-ratio methods),interpolation methods(geostatistical methods,regression models,and machine learning models),validation methods(probability sampling,data splitting,and cross-validation) and indices of accuracy assessments in soil PSF interpolation and soil texture classification(rank correlation coefficient,mean error,root mean square error,mean absolute error,coefficient of determination,Aitchison distance,standardized residual sum of squares,overall accuracy,Kappa coefficient,and Precision-Recall curve) and uncertainty analysis indices(prediction and confidence intervals,standard deviation,and confusion index).Moreover,we summarized several paths on improving the accuracy of soil PSF interpolation,such as improving data distribution through effective data transformation,choosing appropriate prediction methods according to the data distribution,combining auxiliary variables to improve mapping accuracy and distribution rationality,improving interpolation accuracy using hybrid models,and developing multi-component joint models.In the future,we should pay more attention to the principles and mechanisms of data transformation,joint simulation models and high accuracy surface modeling methods for multi-components,as well as the combination of soil particle size curves with stochastic simulations.We proposed a clear framework for improving the performance of the prediction methods for soil PSFs,which can be referenced by other researchers in digital soil sciences.