**2.3 Current applications of remote sensing in epidemiology**

Current research is focusing on the capabilities of remote sensing and GIS data to perform various spatial analyses. The goal is to describe the landscape, or land cover, composition that explains better the distribution of diseases. In this regard spatial models mounted on different software platforms are being developed and validated with observed data to predict spatial phenomenon and in particular diseases that are of public concern. The approaches used to process, analyze and fit data into models are also constantly evolving as new software programs and tool extensions and functionalities become available. There is also an increase in exploratory research involving mathematical and statistical models which aim to capture both the deterministic and the stochastic components during data analyses. In such cases dynamic models involving Susceptible Infectious and Recovered models (SIR) used to model highly infectious diseases such as COVID-19 are being extended by adding probabilistic based components in order to model uncertainty in the behavior of the diseases among the affected population. In these instances, environmentally determined infectious diseases rely on climatic and weather data derived from remote sensing for effective modeling. Furthermore, Satellite imagery could be used to analyze the socioeconomic changes that are currently taking place as a result of COVID-19. This could also be useful in identifying the impacts that measurers such as the national lockdowns are having on the environment. Depending on the resolution of the RS data product being utilized, frequency of data capture and timeliness of the data capture, there is a high potential for remote sensing to be used a tool for deployment and monitoring the effects of health interventions implemented to fight COVID-19.

Point process models such as Log-Gaussian Cox Process had been developed are also being used to model climatic and environmental data on fine geographic scale. These models combine a Poisson process in the first level with a Gaussian Process at the second level and are used to analyze point patterns. In such cases very highresolution remotely sensed data would be used to enhance boundary delineation during mapping especially at local scales. Tools like spatial scan statistics had also been used to identify and map disease clusters and to determine the key driving factors resulting in the identified clusters. Spatial scan statistics defined as the maximum likelihood ratio statistics over a collection of scanning windows had also been widely used to determine clustering in space and time. Research studies applying remote sensing in disease mapping and epidemiology are currently being undertaken at various degrees of complexity as new methods and techniques become available. The primary focus had been on the use of remote sensing and GIS capabilities to quantify various disease determining factors and to estimate the probability that vector-borne diseases will be more abundant in some of the identified habitats as well as to determine the factors necessary for transmission, survival and reproduction.
