Interpolation Of Climatic Resources In Multi-core Research On Parallel Computing Environment

Climate resources are most closely related to the agricultural production, andgreatly influence the growth of major crops, which directly affect the majority of theincome of farmers and national food security.Using geographic information systems torealize small grids of climate resource projections and analysis, by detailing theclimate characteristics of spatial distribution of the resources under different te rrain,rational layout of regional characteristics of agricultural development, agriculturalproduction, as well as future plans provide a sound scientific basis for agriculturaldevelopment.In the refinement of the agro-climatic zoning, the need for a large number ofhistorical data,using a variety of small grid interpolated interpolation algorithms, andinterpolation algorithm for massive data processing complexity, requires a lot ofcomputing time, causing the reaction speed of the system is slow and overallperformance of the system.Kriging interpolation method, is a local interpolation method, is an optimalinterpolation method.Kriging interpolation is established on the basis of variation inspatial analysis functions, limited area inside the area of optimal variable values areunbiased estimates of a method.Compared with other interpolation methods, takinginto account not only the plug and adjacent known points of spatial correlation, and togive the estimated error.Using this method in Hunan province small grid point masstemperature data interpolation, because the serial algorithm of Kriging is morecomplex, and circulating more cause for compute-intensive, and small number of gridpoints reached107result calculation volume, long calculation times.Based on atraditional single processor mode, when working with large-scale massive data,because the calculation time is too long, unable to meet the needs of real-timeanalysis.Resolved in the high-performance computers or distributed computer, butunable to meet the needs of real-time computing on the PC.Multi-core parallelcomputing technology is one of the important directions of technology development.Multi-core computing is the use of parallel processing technology, developmentof parallel, simultaneous execution of multiple tasks, and provide an ideal platform toimprove the multi-core processor performance, but also one of the key technologiesfor improving system performance, so that the current level of computer processing ofa qualitative leap.Therefore, the use of multi-threaded programming techniques and the OpenMP programming model based on the shared storage, the serial algorithm ofKriging for improved, not only improves the efficiency of the algorithm, and a largenumber of experimental data also shows that the high cost of the improved Krigingalgorithm to satisfy the demand for Hunan area network with real-time interpolation.Using a multi-core parallel computing technologies to build the informationsystem, through multithreading programming techniques, making parallel computinghardware resources to really play a role in the development process of informationsystem, full utilization of hardware resources, greatly improve the reaction speed andoverall system performance.