FLOODMAL

Malaria is a vector-borne disease which occurs where Anopheles mosquito vectors, Plasmodium parasites, and vulnerable human populations coincide. There are many opportunities to break the malaria transmission cycle; considerable recent progress has been achieved by protecting human populations using insecticide-treated bednets and house-spraying, and treating the parasite using drug therapies when prevention fails. Nonetheless, these gains are threatened by emerging resistance of mosquitoes to insecticides and drug resistance in the parasite. In addition to protecting human hosts and treating the parasite, however, there is an additional opportunity: control of mosquito vector populations - not only within houses, but in aquatic habitats in the wider environment.

In order to target mosquitoes in these larval habitats, we must understand when and where they occur. Whilst habitats can be identified from satellite imagery, this approach can only provide a snapshot of their current distribution; in order to implement interventions such as larviciding or habitat management, we need to know their distribution in advance. Prediction in floodplain environments can only be possible through an understanding of how water bodies are created by flooding, and how these water bodies are used by mosquito vector populations. It is therefore necessary to combine the respective scientific disciplines of hydrology and ecology to gain the process-based understanding needed to predict larval habitat distributions for targeting.

Numerous flood modelling techniques have been used for decades to predict the distribution of water across a floodplain, yet this approach has not been used to predict aquatic mosquito habitats. Our project will combine an established flood model with an agent-based model which simulates the feeding, breeding and dispersal of a mosquito population at landscape scale, based on the interactions between these malaria vectors and the distributions of their habitat and hosts. We will integrate these established modelling approaches with innovative use of latest generation earth observation imagery and field data from the Barotse floodplain of the Zambezi River in Zambia's Western Province. We believe that this interdisciplinary approach will allow us to predict the distribution of highly seasonal hotspots of malaria vector larval abundance for the first time at an appropriate scale and extent for implementing interventions.

This platform will not only enable us to predict the short-term dynamics of malaria vector hotspots, but also to assess the potential impact of future climate change on malaria transmission hazard between now and the end of the 21st century. Whilst the effect of temperature on malaria vectors has been widely studied, the impacts of changed river flows on malaria hazard - and the intervention efforts which will be required to address them - remain largely unknown. To provide the resilience required by the Zambian government's campaign to eliminate malaria by 2020, this is a knowledge gap which must be filled.

Who will benefit from this research?
1) Flood modellers
2) Earth observation practitioners
3) Spatial ecologists
4) (Micro-)Epidemiologists
5) Health practitioners and agencies (e.g. Western Province Medical Office (WPMO), Zambia, Canadian Coalition for Global Health Research, (CCGHR))
6) Malaria intervention campaigns -Zambian government (National Malaria Control Centre, NMCC), Global Malaria Elimination campaign
7) Innovative Vector Control Consortium (IVCC)
8) General public

How will they benefit from this research?
1) Gain published methodologies for application of new technology/data sources (Pleiades tri-stereo imagery, UAV hardware and processing techniques, commercial sonar for bathymetry) to provide inputs for established hydraulic flood models; methodology for validation of model outputs using earth observation.
2) Potential to use our suite of radar-processing techniques to map floodwaters routinely from freely available Sentinel-1 data.
3) Improvement of existing agent-based models offered by the development and calibration of an agent-based model employing a 2-scale approach for simulating vector populations for the first time over a large area at fine resolution.
4) Provision of detailed larval abundance maps from processed-based understanding - this has been the missing input for transmission models often previously reliant on static habitat maps or assumed homogeneity of vector populations.
5) The prediction of hotspots of transmission hazard, both now and in the future will be used to inform resource distribution, prioritise staff deployments and develop the required health systems capacity to facilitate resilience in the face of a dynamic challenge.
6) The ability to obtain larval abundance hotspot maps with sufficient predictive power and local resolution will allow the unprecedented opportunity to judiciously target larval source management campaigns, and the FLOODMAL approach will be applicable to the large contiguous area over which such interventions must be deployed.
7) Detailed information on the dynamics of malaria vector hotspots will elucidate the required characteristics (e.g. longevity, means of application) of the latest generation of vector control methods.
8) Improved estimates will be provided of the likely impact of climate change on the magnitude and seasonality of malaria transmission hazard, and the targeted interventions which will be permissible for the first time due to this invaluable new information will contribute significantly to the ongoing campaign to eliminate malaria.

We will publish widely in a range of journals to disseminate methodologies and results to academics, whilst concerted engagement with relevant stakeholders (WPMO, CCGHR, IVCC, NMCC) will ensure that research outputs can be used by health services/agencies and especially by intervention campaigns. We believe that the FLOODMAL project's potential to enable targeting of larval source management offers a significant opportunity to drive forward the existing campaign to eliminate malaria from Zambia by 2020 as part of the wider global elimination program. The project will facilitate improved resource targeting and increased resilience of malaria control campaigns to climate change impacts and therefore contribute to a reduction of the burden of morbidity, mortality and reduced economic prosperity which is imposed by malaria.