Coupling of a Water Container Energy Balance Model with Gridded NASA Earth Science Products for Dengue Risk Applications

Dengue poses a risk to 3.5 billion people in over 100 countries across the tropical and sub-tropical Americas and Africa, Southeast Asia, India, and Oceana.  Dengue infections are estimated to total nearly 400 million per year worldwide, with about 1% of cases exhibiting the severe and often-deadly dengue hemorrhagic fever.  Both the geographic range and the magnitude of dengue infections have increased in the past 50 years, due to population growth and urbanization in endemic areas, increases in global mobility and trade, and the discontinuation of insecticide spraying programs (particularly in the Americas) because of financial and environmental concerns.  Dengue is currently present in portions of southern Florida and along the U.S./Mexico border region.

The primary dengue vector mosquito Aedes aegypti is closely associated with humans.  It lives exclusively in urban and semi-urban areas, preferentially bites humans, and spends its developmental stages in artificial water containers.  The primary limiting climatic factor for Ae. aegypti survival is cold temperature, with an approximate lower boundary of 10°C average winter temperature.  Water availability is a limiting factor in arid regions, and the seasonality of Ae. aegpyti depends on temperature in subtropical regions and rainy seasons (water availability) in tropical regions.  Climate effects extend to development of Ae. aegypti immature mosquitoes in artificial containers.  Potential containers for Ae. aegypti immature development include, but are not limited to, small sundry items (e.g., bottles, cans, plastic containers), buckets, tires, barrels, tanks, and cisterns.  Successful development of immature mosquitoes from eggs to larvae, pupae, and eventually adults is largely dependent on the availability of water and the thermal properties of the water in the containers. 

Modeling of Ae. aegypti populations is often done using dynamic life cycle models that simulate the life cycles of cohorts of mosquitoes using a mechanistic approach.  Perhaps the greatest limitation of these dynamic life cycle models is the continued use of simplistic empirical relationships to predict water temperature in and water loss from containers based on several meteorological variables.  Recent work has shown that physics-based approaches toward modeling container water properties are promising for resolving the complexities of container water dynamics.  Such models solve for the energy balance of the water inside of the container, taking into account shortwave (solar) and longwave (terrestrial) radiation and heat fluxes (sensible, latent, and ground).  The methodology is similar to what is done in land surface models to simulate ground surface and soil temperature, except modified for container dimensions.  An energy balance container model developed by the author, termed the Water Height And Temperature in Container Habitats Energy Model (WHATCH'EM), solves for water temperature and height for user-specified containers with readily available weather data.  Realistic estimation of water temperature and height from WHATCH'EM has potential to improve output from mosquito population models.   

WHATCH'EM was originally developed to be run for user-specified containers at point locations where sufficient input meteorological data is available.  Here we introduce modifications to WHATCH'EM so that the model can be run using gridded NASA Earth Science datasets.  This allows for the model to be used for dengue risk applications in areas with insufficient or unrepresentative in situ meteorological data.   Specific NASA Earth Science products used as input to WHATCH'EM include the Modern Era Retrospective-Analysis for Research and Applications (MERRA), the Global Land Data Assimilation System (GLDAS), the North American Land Data Assimilation System (NLDAS), and the Tropical Rainfall Measuring Mission (TRMM).  Variables from these datasets input into WHATCH'EM include near-surface air temperature, relative humidity, rainfall, downward shortwave and longwave radiation, near-surface wind speed, and soil temperature. 

Results from the WHATCH'EM simulations using NASA Earth Science products (WHATCH'EM-NASA) are validated against several datasets across varying climates.  Direct container measurements of water temperature and water height were conducted for one-month periods in summer 2013 at Boulder CO, Veracruz MX, and Orizaba MX.  Additional data points are also available from a household survey campaign in Hermosillo MX in summer 2014 and from various published studies.  The WHATCH'EM-NASA simulations are compared and contrasted with both direct container observations and with WHATCH'EM simulations run using in situ meteorological observations to best characterize biases in the WHATCH'EM-NASA simulations.  Overall validation results are discussed, along with Ae. aegypti habitat suitability and seasonality differences between the humid tropical and desert Mexican locations analyzed, and plans to couple WHATCH'EM-NASA with mosquito life cycle models.