Research about Beale Air Force Base

USGS2021

Experiences in LP-IoT: EnviSense deployment of remotely reprogrammable environmental sensors

The advent of Low Power Wide Area Networks (LPWAN) has improved the feasibility of wireless sensor networks for environmental sensing across wide areas. We have built EnviSense, an ultra-low power environmental sensing system, and deployed over a dozen of them across two locations in Northern California for hydrological monitoring applications with the U.S. Geological Survey (USGS). This paper details our experiences with the design and implementation of this system across two years, including six months of continuous measurement in the field. We describe the lessons learned for deployment planning, remote device management and programming, and system co-design with a domain-expert from the USGS.

California↗

USGS2011

Using avian radar to examine relationships among avian activity, bird strikes, and meteorological factors

Radar systems designed to detect avian activity at airfields are useful in understanding factors that influence the risk of bird and aircraft collisions (bird strikes). We used an avian radar system to measure avian activity at Beale Air Force Base, California, USA, during 2008 and 2009. We conducted a 2-part analysis to examine relationships among avian activity, bird strikes, and meteorological and time-dependent factors. We found that avian activity around the airfield was greater at times when bird strikes occurred than on average using a permutation resampling technique. Second, we developed generalized linear mixed models of an avian activity index (AAI). Variation in AAI was first explained by seasons that were based on average migration dates of birds at the study area. We then modeled AAI by those seasons to further explain variation by meteorological factors and daily light levels within a 24- hour period. In general, avian activity increased with decreased temperature, wind, visibility, precipitation, and increased humidity and cloud cover. These effects differed by season. For example, during the spring bird migration period, most avian activity occurred before sunrise at twilight hours on clear days with low winds, whereas during fall migration, substantial activity occurred after sunrise, and birds generally were more active at lower temperatures. We report parameter estimates (i.e., constants and coefficients) averaged across models and a relatively simple calculation for safety officers and wildlife managers to predict AAI and the relative risk of bird strike based on time, date, and meteorological values. We validated model predictability and assessed model fi t. These analyses will be useful for general inference of avian activity and risk assessment efforts. Further investigation and ongoing data collection will refine these inference models and improve our understanding of factors that influence avian activity, which is necessary to inform management decisions aimed at reducing risk of bird strikes.

California↗