headerphoto

Malaria Machine Learnign for interventions

IAEAECMWF A project in collaboration with IAEA running 2021-2023, we are applyign machine learning algorithms to calculate the most cost effective way of implementating the sterale insect technique releases that also account for seasonality and interannual climate variability. A new parameterization for SIT has been inserted in 2022 and the machine learning algorithms are currently under investigation.



Malaria Early Warning System

ECMWFECMWF monthly and seasonal forecasts are being used to drive VECTRI to create a first prototype pan-Africa operational forecasting system for malaria. The system has been validated with colleagues from the ministries of health in Malawi, Uganda and Rwanda. The pilot forecast system runs for the whole of the African continent. Please contact the authors if you are interested in receiving hindcast data as part of this validation exercise.


ISIMIP

ISIMIPISIMIP is a malaria model intercomparison project which contributes to the impacts section of the next IPCC 5th assessment report. VECTRI is presently the only dynamical model contributing to the project that runs on a daily timestep and thus accounts for subseasonal variations in climate


HEALTHYFUTURES

Healthy FuturesHealthy Futures is an EUFP7 project running from 2011-2014. Its focus is on the impact of environment, including climate change, on malaria transmission in Eastern Africa. Led by Prof. David Taylor of TCD in Dublin (now at Singapore national university) the project is supported by institutes throughout Europe and Eastern Africa. According to the HF website: @quot the project aims to construct a disease risk mapping system for three water-related, high-impact VBDs (malaria, Rift Valley fever and schistosomiasis) in eastern Africa, taking into account environmental/climatic trends and changes in socio-economic conditions to predict future risk @quot


QWeCI

QWeCIQuantifying weather and climate impacts on health in developing countries (QWeCI) is a project led by Dr. Andy Morse of the University of Liverpool and which was co-designed by Adrian Tompkins here at ICTP. While the project also examines the variation of malaria transmission over longer timescales, the main focus is on the predictability of malaria on monthly to seasonal timescales. The project has equal number of partners in Europe and Africa, and initiated the collaboration that has led to the ECMWF-VECTRI pilot forecasting system that is currently under evaluation.

Links

The VECTRI authors would also like to ackowledge the following resources/projects, which have been used in the VECTRI model or for its evaluation.

MAP : Gridded assessment of PR and EIR values using a Bayesian model to incorporate survey data.

Afripop : high resolution population density data for Africa.

GRUMP: global population data on a high spatial resolution.

NEWS

20/12/2021: VECTRI v1.9 released!!!
NEW features in VECTRI v1.9
This upgrade to the code mostly focuses on improvements to the hydrological scheme. The manual will be upgraded soon to outline these in more detail.
  • Ability to read in externally generated pond fraction from a full hydrological modelling scheme
  • New map of permanent breeding sites using pond edge fraction derived from Sentinal 2 data (20m/150m resolution Africa/global, aggregated to 5km)
  • Correction and relaunch of the revised Asare pond scheme which allows for non-linear runoff and overflow and is validated against in situ observations and high resolution modelling in Ghana and Niger (see Asare et al. 2016a/b)
  • Maps of soil texture used to allow spatially varying infiltration rates according to clay/silt/sand proportions
  • New -v command line option to pass modified paramters to code (avoiding the need to place options in the vectri.options file).

  • Version 1.8.0/1.8.1
    Jan 2020: Diffusion of vectors between gridcells added.
    May 2019, Version 1.7.0 has now been released (May 2019) which introduces a massively simplified interface to make it much easier to run the code with gridded data. The model no longer accepts text file input, only netcdf input is possible (a grib interface is currently under testing). A new (and slowly improving) manual is also available.

    Did you know?

    VECTRI is now coupled to seamless monthly-to-seasonal forecasts from ECMWF to create a pilot pan-Africa operational forecasting system for malaria.

    read more>>
    VECTRI downloads: