Weather surveillance radar as an objective tool for monitoring bat phenology and biogeography

: Bats provide a wealth of vital services to humans, such as pollinating plants and managing populations of insects that act as agricultural pests and disease vectors. Despite their direct importance, it is inherently difficult to monitor bats across large spatiotemporal scales, in part due to their nocturnal activity and long-distance, high-altitude flights. Networks of weather surveillance radars provide continuous measurements of the airspace at continental scales, revealing the abundance, distribution, behaviour, and phenology of bats aloft. Using a network of polarimetric Doppler weather radars in the USA, we demonstrate applications of this technology to monitor large bat colonies, highlight recent discoveries made using radar surveillance of bats, and discuss future prospects for extending these techniques as part of a standardised system of global bat monitoring infrastructure.


Introduction
Bats are effective bioindicators of environmental conditions and provide a host of ecosystem services that humans rely on in agriculture and public health [1]. Bats also face a number of threats including landscape changes, loss of habitat, human encroachment, and spread infectious diseases [2]. With billions of dollars at stake in the USA alone [3], there is a growing imperative to monitor and assess bat populations so that conservation efforts can be applied efficiently and effectively. While the need for these observations exists, practical difficulties in obtaining standardised measurements of volant taxa have made developing objective surveillance techniques and protocols-especially over large spatiotemporal scales-an ongoing challenge [4].
The presence of bats on radar displays has been documented since the late 1960s, and the capability of using remote measurements of bat colonies for ecological applications was immediately recognised [5]. Despite the obvious potential, it was nearly 40 years later until the first long-term study of bat abundance and biogeography was performed using the US national network of weather surveillance radars (NEXRAD) in 2008 [6]. Since this pioneering study by Horn and Kunz [6], our fundamental understanding of these data has continued to mature, and analytical methods for extracting their underlying biological information have progressed [7]. In the following sections, we provide a brief overview of weather radar as it applies to bat observations, review recent research findings that have been made possible through radar observations, and discuss the potential of applying these techniques over large spatial (continental) and temporal (decadal) scales to obtain a general understanding of the effects of global change on bat colonies.

Weather radar theory and methods
The US network of Doppler weather surveillance radar operates at S-band and routinely detects airborne animals in flight, making it ideal infrastructure for monitoring the ecology of the lower atmosphere (i.e. aeroecology) [8]. The network has been fully upgraded to dual-polarisation technology, allowing new possibilities for identifying and discriminating taxa aloft by their radar scattering characteristics [9]. Moreover, advances in electromagnetic scattering [10] and radar simulation [11] have enabled new methods for quantifying the number of organisms in the airspace [12]. By applying the suite of these techniques, we have enabled long-term analyses of the radar data archive to explore the ecology of bats.

Results of recent research
Weather radar has the capability of revealing the biogeography of bat colonies as they emerge into the airspace (Fig. 1), enabling quantitative mapping of population distributions. The ability to map these distributions over time will be especially important for determining the risks associated with anthropogenic and natural hazards. One example is determining the effect of severe weather events on bat mortality and habitat destruction, such as the case of Hurricane Harvey in August of 2017 (Fig. 2). In these instances, impacts on bat colonies are acute, but the overall long-term impacts on populations are not well-understood. Radar observations have the capability of providing surveys of these effects and their persistence in time.
Outside of extreme events, the persistent hazards of global change threaten to harm bat populations, further motivating radar surveillance. A previous study used 11 years of radar measurements to show that bats at five colonies exhibited behavioural responses to changes in weather and climate, engaging in riskier behaviour (i.e. emerging in brighter conditions) when experiencing drought [13]. In recent work, we found that a large bat colony has been steadily changing its migration phenology over decadal timescales, while increasing the number overwinter individuals at the site [14]. The ability of radar to provide nightly estimates of bat abundance allows observations across the full phenological cycle over decadal time spans (Fig. 3). Moreover, these rich datasets can be combined with corresponding meteorological and environmental observations to deduce the correlates of bat abundance, distribution, migration, and phenology. In such cases, routine radar surveillance can yield a better understanding of the ecological implications of global change, but more work is still needed to determine the spatial extent and wider impact of these changes.

Global prospects
Although much of the existing effort on weather radar surveillance of bat populations has focused on the NEXRAD network in the USA, weather radar infrastructure is widely considered a global imperative and much of the Earth's landmass is overlooked by weather radar (Fig. 4). Systems of shared calibration standards and methods have been developed by the weather radar community, and groups such as the World Meteorological Organisation continue to provide benchmark infrastructure goals for individual nations. As the value of weather radar in ecological applications continues to be realised, a growing multidisciplinary, multinational community of physical scientists, engineers, and biologists are developing the capacity to use these methods to benefit a wide breadth of stakeholders [15]. Strengthening research partnerships, cooperative data-sharing policies, and data quality standards hold the key to applying the global suite of radar infrastructure to some of the most pressing questions on large-scale influences of global change on bat populations. Furthermore, as remote-sensing continues to play a greater role in ecological and environmental studies, our capacity to observe animal systems at comparable spatiotemporal scales will motivate large-scale adoption of radar methods.

Conclusion
Despite being deployed with the mission of weather surveillance, NEXRAD has provided new information on the ecology, biogeography, and phenology of bat populations, revealing new and unknown information that would be difficult-if not fully impractical-to obtain through other methods. While radar certainly cannot replace field surveys and biologging for some applications, its capacity for providing standardised observations makes it an exciting prospect for inter-continental comparisons. Indeed, the global nature of weather radar infrastructure, coupled with common calibration standards and analytical methods, could yield a truly global view of bat ecology in a changing world.

Acknowledgements
P. Stepanian received funding for this work through a Marshall Sherfield Fellowship. We would like to thank the Turkish State Meteorological Service for maintaining and providing the WMO global radar site location database.