Precipitation Measurement: Gauge Deployment

Abstract

The deployment of a network of rain gauges involves two stages: first, the broad distribution of gauges across a region or catchment and, second, the specific location of each gauge. The spatial distribution of gauges across a region will be determined by a combination of: (a) physical characteristics of the region, including the topography and the natural variability of precipitation, and (b) institutional factors, such as the purpose of the network (setting the accuracy and level of detail of data required as well as the resources available). The specific location of gauges will be determined by purely local factors, which would affect the accuracy of the measurements. These are primarily concerned with site exposure and wind effects that could lead to errors in catch and make the data unrepresentative of a larger area.

Multicriteria design of rain gauge networks for flash flood prediction in semiarid catchments with complex terrain

“…Despite the availability of weather radar data at high spatiotemporal resolution, ground‐based rain gauge networks are still necessary to enable accurate estimation of storm rainfall input to catchments during flash flood events. Unlike radar systems which (indirectly) measure continuously in space, a network of rain gauges represents a discrete finite sampling of the two‐dimensional pattern of land surface precipitation depths [ Robinson , 2005]. Even though rain gauges provide relatively accurate (direct) rainfall measurements [ Villarini et al , 2008] they are only representative of a limited spatial extent and must therefore be interpolated to obtain estimates of the areal pattern of precipitation or mean areal precipitation (MAP) across a catchment [e.g., Villarini et al , 2008; Garcia et al , 2008].…”

“…The accuracy of gauge‐based areal estimates of rainfall depends on both the number of gauges in the network, and on their spatial location and distribution. Specification of these characteristics of a rain gauge network is necessarily related to the natural variability of the precipitation across an area [ Rodríguez‐Iturbe and Mejía , 1974], which arises from physiographic and hydrometeorological factors [ Eagleson , 1967; Rodríguez‐Iturbe and Mejía , 1974; Robinson , 2005] including topography and climate [ Harris et al , 1996; Pandey et al , 1999]. It is typically assumed that in semiarid mountainous areas, having significant variability and intermittency [ Barancourt et al , 1992] larger numbers of rain gauges are needed to characterize and monitor precipitation fields than in flat terrain and/or regions where rainfall results from frontal storm systems [ Robinson , 2005].…”