Speed-up Techniques for High-Resolution Grid Data Processing in the Early Warning System for Agrometeorological Disaster

농업기상재해 조기경보시스템에서의 고해상도 격자형 자료의 처리 속도 향상 기법

  • Park, J.H. (R&D Center, EPINET Co., Ltd.) ;
  • Shin, Y.S. (R&D Center, EPINET Co., Ltd.) ;
  • Kim, S.K. (R&D Center, EPINET Co., Ltd.) ;
  • Kang, W.S. (R&D Center, EPINET Co., Ltd.) ;
  • Han, Y.K. (R&D Center, EPINET Co., Ltd.) ;
  • Kim, J.H. (National Center for Agro-Meteorology, Seoul National University) ;
  • Kim, D.J. (National Center for Agro-Meteorology, Seoul National University) ;
  • Kim, S.O. (National Center for Agro-Meteorology, Seoul National University) ;
  • Shim, K.M. (National Institute of Agricultural Sciences, RDA) ;
  • Park, E.W. (Department of Agricultural Biotechnology, Seoul National University)
  • Received : 2017.08.07
  • Accepted : 2017.09.18
  • Published : 2017.09.30


The objective of this study is to enhance the model's speed of estimating weather variables (e.g., minimum/maximum temperature, sunshine hour, PRISM (Parameter-elevation Regression on Independent Slopes Model) based precipitation), which are applied to the Agrometeorological Early Warning System ( The current process of weather estimation is operated on high-performance multi-core CPUs that have 8 physical cores and 16 logical threads. Nonetheless, the server is not even dedicated to the handling of a single county, indicating that very high overhead is involved in calculating the 10 counties of the Seomjin River Basin. In order to reduce such overhead, several cache and parallelization techniques were used to measure the performance and to check the applicability. Results are as follows: (1) for simple calculations such as Growing Degree Days accumulation, the time required for Input and Output (I/O) is significantly greater than that for calculation, suggesting the need of a technique which reduces disk I/O bottlenecks; (2) when there are many I/O, it is advantageous to distribute them on several servers. However, each server must have a cache for input data so that it does not compete for the same resource; and (3) GPU-based parallel processing method is most suitable for models such as PRISM with large computation loads.


Supported by : 국립농업과학원


  1. Ahn, J. B., J. N. Hur, and A. Y. Lim, 2014: Estimation of fine-scale daily temperature with 30 m-resolution using PRISM. Atmosphere 24(1), 101-110.
  2. Amdahl, G. M., 1967: Validity of the single processor approach to achieving large-scale computing capabilities. AFIPS Conference Proceedings 30, 483-485.
  3. Chung, U., K. Yun, K. S. Cho, J. H. Yi, and J. I. Yun, 2009: The PRISM-based rainfall mapping at an enhanced gridcell resolution in complex terrain. Korean Journal of Agricultural and Forest Meteorology 11, 72-78.
  4. Kim, C. Y., Y. C. Kim, Y. C. Kim, S. M. Lee, Y. K. Kim, and D. W. Seo, 2014: Performance enhancement of distributed file system as virtual desktop storage using client side SSD cache. Computer and Communication System 3(12), 433-442.
  5. Kim, D. J., and J. I. Yun, 2013: Improving usage of the Korea meteorological administration's digital forecasts in agriculture: 2. Refining the distribution of precipitation amount. Korean Journal of Agricultural and Forest Meteorology 15(3), 171-177.
  6. Mielikainen, J., B. Huang, and H. L. A. Huang, 2013: Speeding up the computation of WRF double-moment 6-class microphysics scheme with GPU. Journal of Atmospheric and Oceanic Technology 30(12), 2896-2906.
  7. Ozimek, I., A. Hrovat, A. Vilhar, and T. Javornik, 2015: GPU computation acceleration of GRASS GIS modules for predicting radio-propagation. International Journal of Communication 9, 76-85.
  8. Rodgers, D. P., 1985: Improvements in multiprocessor system design. ACM SIGARCH Computer Architecture News archive 13(3), 225-231.
  9. Sugiura, T., 1999: Prediction of full bloom date of pear using air temperature. Agriculture and Horticulture 54(10), 146-149.
  10. Zhang, J., S. You, and L. Gruenwald, 2015: Efficient parallel zonal statistics on large-scale global biodiversity data on GPUs. Proceedings of the 4th International ACM SIGSPATIAL Workshop on Analytics for Big Geospatial Data, 35-44.