Hydrologic Modeling and System Optimization for IoT Flood Management

Abstract

The increasing frequency and severity of storms due to climate change is magnifying flooding impacts. The Internet of Things (IoT) revolution promises more ubiquitous sensing capabilities. When applied to water systems, IoT has the potential to increase insights into how hydrologic systems respond to extreme rainfall events, aiding in emergency management efforts before and during extreme weather events. In this paper, we provide a way to translate forecasted extreme rainfall events into flood impacts and optimize an IoT sensor network for real-time flood monitoring. First, we created a hydrologic model for a study area: the Dell Pond watershed in Charlottesville, Virginia. We used ArcGIS to obtain parameters for the model from geospatial datasets such as elevation, soils, land use, and land cover. The parameters obtained from ArcGIS, alongside the National Oceanic and Atmospheric Administration (NOAA) rainfall precipitation data, and readings from the IoT water sensors were combined to create a hydrologic model in HEC-HMS. To optimize the IoT sensor monitoring network and explore systems integration of the model and sensors, we first created models to determine the battery life of a sensor in the network, since the IoT sensors are battery powered with no additional power harvesting capability. We also deployed a new water level and a soil moisture sensor using the IoT network for the study watershed. The methods for estimating the battery life of the IoT sensor and the prototype deployment can be built on in future research to advance next-generation flood management systems that integrate computational models and IoT monitoring networks.