Infectious Disease Prediction

Based on seasonal climate prediction information, observation data, and malaria incidence data, we predict malaria incidence in South Africa for the next nine months using machine learning models. In the future, we plan to expand our prediction to longer periods of time, to other regions, and to other infectious diseases.

・APL Virtualearth
・Malaria prediction using Machine learning models

VAi aims to contribute to the resolution of policy issues and the development of a sustainable socio-economic system by creating and disseminating information that meets various needs.

• The infectious disease predictions are generated using the malaria prediction model, based on the dynamically downscaled SINTEX-F climate predictions, and malaria incidence data, are provided to the iDEWS bureau (South African epidemic disease laboratory, weather service, and provincial health officials). We plan to expand to other regions and other infectious diseases such as Mozambique, Kenya, Ghana, India, Bangladesh, etc. in the future.

NAMR is Numerical Analysis Method Repository. (Numerical analysis methods: artificial intelligence, sparse modeling, finite element method, particle method, verification method)

• SINTEX-F seasonal prediction system

  The ocean-atmosphere coupled model provides climate prediction for several seasons to two years into the future.

• Climatic downscaling system

  Climate prediction data is refined by the regional atmospheric model, the Weather Research and Forecasting (WRF) system.

  

  Climatic downscaling system

• Automatic fishing point detection from AIS position data

  Tool for fishing point detection derived from AIS position data of fishing vessels

• Virtual fish farming cage

  High-performance estimation by integration of numerical model that reproduces fish farming cage and echo data of fish finder

  

  Reproduction of yellowtail distribution by virtual fish farming cage (upper) and the corresponding fish finder echo simulation (lower)

4DVE develops and maintains a system that allows users to map variety of analysis data including time fluctuations, and seamlessly combine and extract information. It is called the 4th Dimensional Virtual Earth.

• FORA（Four-dimensional Variational Ocean ReAnalysis for the Western North Pacific）

  Daily ocean environment product at 10km resolution.
  
  ・FORA（Four-dimensional Variational Ocean ReAnalysis for the Western North Pacific)
  

• FORP

  Future prediction data set of ocean environments using atmospheric data from multiple climate models of CMIP5 as an external forcing, with a 10-km grid for the North Pacific and a 2-km grid for the sea around Japan.
  
  ・FORP（Future Ocean Regional Projection dataset for coastal applications in Japan）
  ・FORP（Future Ocean Regional Projection dataset for the Western North Pacific）

  

  Changes in sea surface temperature from the present climate in the waters near Japan by FORP

• Lower trophic level model products for marine ecosystem

  A prediction data set of lower-trophic level marine ecosystems (zooplankton, etc.) by coupling online with FORA and FORP ocean models

• Estimation of marine species distribution

  Species distribution data estimated/predicted by applying ocean environment data to potential fishing ground estimation model

• Biomass estimation

  Estimated growth and amount data of target fishes by using virtual fish farming cage