The potential of remote sensing for improved infectious disease ecology research and practice

Abstract

Outbreaks of COVID-19 in humans, Dutch elm disease in forests, and highly pathogenic avian influenza in wild birds and poultry highlight the disruptive impacts of infectious diseases on public health, ecosystems and economies. Infectious disease dynamics often depend on environmental conditions that drive occurrence, transmission and outbreaks. Remote sensing can contribute to infectious disease research and management by providing standardized environmental data across broad spatial and temporal extents, often at no cost to the user. Here, we (i) conduct a review of primary literature to quantify current uses of remote sensing in disease ecology; and (ii) synthesize qualitative information to identify opportunities for further integration of remote sensing into disease ecology. We identify that modern advances in airborne remote sensing are enabling early detection of forest pathogens and that satellite data are most commonly used to study geographically widespread human diseases. Opportunities remain for increased use of data products that characterize vegetation, surface water and soil; provide data at high spatio-temporal and spectral resolutions; and quantify uncertainty in measurements. Additionally, combining remote sensing with animal telemetry can support decision-making for disease management by providing insights into wildlife disease dynamics. Integrating these opportunities will advance both research and management of infectious diseases.

Keywords: infectious disease ecology; pathogens; remote sensing; soil moisture; surface water; vegetation.

Figure 1.

Use of remote sensing variables across diseases with different host types. The y-axis measures proportion of studies of diseases with each definitive host type that used at least one variable of the focal type (e.g. approx. 51% of human disease studies used at least one vegetation index). Note the absence of precipitation and human population data in studies of plant diseases and the over-representation of plant studies in the vegetation properties category. The eight most common variable types are shown and variables are grouped thematically. Numbers in the legend indicate the total number of studies with each definitive host type; studies are counted multiple times if they study a disease with multiple definitive host types (e.g. zoonoses).

Figure 2.

Study effort across host types. (a) Number of diseases with each definitive host type. Diseases with multiple definitive hosts can appear in multiple columns. Numbers differ from figure 1 because they count unique diseases, not studies. Some diseases of humans (grey) have humans as accidental hosts, not definitive hosts. Most diseases in all categories except plant diseases are zoonotic, i.e. infecting both wild animals and humans. (b) Study effort for individual host taxa. This metric considers the focal host(s) of a study rather than the definitive host of a disease. Studies are counted multiple times if they use data from multiple hosts.

Figure 3.

Spatio-temporal scales of key variables for infectious disease ecology. (a) Spatial and temporal resolutions of currently available remote sensors. Rectangles represent the range of spatial and temporal resolutions currently available from each sensor type. Resolutions can be made coarser, so rectangles represent the finest available resolutions (i.e. rectangles extend up and right). Note that different sensor types provide different variables. (b-f) Key variables for the study of five focal infectious diseases. Points show the ideal spatial and temporal resolution of each variable for the study of the focal disease. Points below or to the left of a coloured rectangle indicate that sufficiently fine resolutions are not yet available from a given remote sensor. For example, in panel (e), soil temperature at a scale relevant to the ecology of soil-transmitted helminths could not be studied using geostationary satellites alone, because spatial resolution is insufficient. Abbreviations: UAV, uncrewed aerial vehicle; DEM, digital elevation model; LiDAR, light detection and ranging; VHR, very high resolution.