WP4 - Coupled Climate-Health Projections

• To integrate a dynamic disease model within a up to seasonal time scales within a seamless ensemble prediction system using monthly and seasonal hindcasts
• To move the integrated modeling system into the experimental decadal ensemble prediction system hindcasts and projections
• To work on the  initial stages of developing a multi agent based modeling system for use  in Senegal

Task 4.1a: Modelling System development  to modify the input output routines from the dynamic Liverpool Malaria Model to work with different ensemble model output   (UNILIV, IC3)

This task will involve the interaction with the seamless model stream production WP3.1 and WP3.2 as well as the downscaling and post processing work in WP 3.1.  The main tasks will be developing the routines to be able to optimally run with a range of data inputs covering a wide range of time scales i.e. monthly to decadal scales.

Task 4.1b: Force malaria specific dynamic models (fully and semi dynamic) with bias corrected monthly and seasonal hindcasts – to develop a seamless system – using verification against control (reanalysis data sets) runs for disease model outcomes  (UNILIV, IC3,  KNUST, CSE, IPD, UNIMA, UP). This is to verify that running the model on a seamless monthly to seasonal ensemble product does not lead to unexpected or incorrect modelling outcomes when compared to reanalysis or seasonal on ensemble prediction system runs.  Further these data will be run pre and post processed via outputs from WP3.1

Task 4.1c: Evaluate skill of malaria and RVF (and tick borne diseases depending on completion of Task 2.1C ) seasonal hindcasts using disease databases (UNILIV, IC3,  KNUST, CSE, IPD, UNIMA, UP, ILRI)

This task takes the validation a stage further than in Task 4.1b where control runs were used.  In this task real world datasets collected within the African partner countries will be used to validate the integrated disease-climate modeling system

Task 4.1d: Force malaria and RVF specific dynamic models with bias corrected decadal hindcasts and projections.  Hindcasts to be verified against control (reanalysis) runs disease model. (UNILIV, IC3)

This task will move the work into an new dimension to run disease models with decadal forecasts using them dynamically to assess the forecasting predictability of decadal forecasting systems.  Here decadal runs can be verified against reanalysis based control runs.  If the integrated models seem to be working within reasonable bounds then forward projection runs will be undertaken with decadal scale forecasts from a range of forecasting systems. The outputs of the ensemble decadal runs will be compared with decadal runs from other climate modeling systems.

Task 4.1e. The initial development of an integrated multi-agency system in Senegal. (UCAD, UNILIV, ILRI)

The task will incorporate model output from the above tasks and combine it with remotely sensed data, GIS layers and stakeholder knowledge.  This system will be developed in conjunction with the development of decision support systems being undertaken in WP 5.1

D4.1.a Report on the seamless  integration of a dynamic disease model with monthly and seasonal ensemble prediction system hindcasts. M24 (UNILIV, IC3,  KNUST, CSE, IPD, UNIMA, UP)

D4.1.b Report on decadal ensemble prediction system integration with a dynamic disease model M30 (UNILIV, IC3)

D4.1c Description of the pilot integrated multi-agent system M30 (UCAD)

D4.1c: Description of the pilot integrated multi-agent system

M4.1a Pilot integration of the existing dynamic malaria model with a seamless monthly and seasonal ensemble prediction system M18 (UNILIV)

M4.1a – Preliminary analysis of the existing dynamic malaria model with a seamle

M4.1b Pilot integration of the existing dynamic malaria model with a decadal ensemble prediction system M24 (UNILIV, IC3)

M4.1b - Pilot Integration of the existing dynamic malaria model with a decadal e