A rainfall monitoring network refers to a network of gauging stations located in space or a region used to measure rainfall. Rainfall monitoring network density is simply the number of rain gauges per area of a specific region. The rain gauge density is an important aspect of monitoring network design, and optimal density refers to the optimal number of rain gauges within a specific region. However, the World Meteorological Organization provides few guidelines for assessing minimum density. One method available for network design is the geostatistical method of rain gauge optimization. Understanding that a monitoring network’s design aims to characterize the spatial variability of the rainfall across a specific region is important. The chapter focuses on the use of rain gauge and radar-based rainfall estimates to design optimal precipitation monitoring networks.
This chapter summarizes principles of radar quantitative precipitation estimates (QPE), as used by the National Weather Service, and considerations in applying radar and gauge QPE to hydrologic modeling tasks. It describes the processing sequence for Weather Service Radar 1988 Doppler (WSR-88D) and gauge-radar precipitation estimates and offers basic principles of radar data acquisition and processing. The chapter provides radar QPE errors’ statistical characteristics and briefly describes basic approaches to statistical blending of radar and rain gauge estimates. The chapter considers the relative value of radar and gauge QPE in situations in which gauge data are readily available, or where radar coverage is compromised. It also summarizes possibilities for using daily rain gauge input to adjust subdaily estimates and lists current sources for WSR-88D, gauge-radar multisensor, and rain gauge data.
This chapter presents some examples of radar rainfall data analyses and applications, exploring how the South Florida management district carries out radar rainfall data estimation. The chapter provides all the steps involved in generating near-real-time data. It discusses radar rainfall data analyses included for rainfall frequency analysis and depth area reduction factors and applications included in near-real-time flood warning system, sewer system modeling, groundwater recharge modeling, and rain gauge network design. Rainfall frequency analyses are often used to generate “design storms” for flood risk assessment. Urban drainage system design and real-time forecasting for multiscale urban drainage systems is complex. The chapter describes a distributed hydrologic model that quantifies hydrologic prediction uncertainty in urban-scale catchments and was used for a groundwater recharge estimate by estimating components of the hydrologic water balance.
This chapter discusses radar rainfall data formats for use in hydrologic models. Radar rainfall estimation offers potential to provide accurate rainfall surfaces because radar can cover entire watersheds completely, provide high spatial resolution, and provide high temporal resolution. Hydrologic models are grouped into two categories: lumped or gridded. A lumped hydrologic model combines all hydrologic processes within a single watershed or sub-watershed into a single element. A gridded model considers hydrologic processes within a rectilinear grid cell. The National Weather Service Weather Service Radar Doppler 88D radars were first installed in the early to mid-1990s. However, the growing national database of radar rainfall estimates represents the first opportunity for hydrologists to conduct in-depth studies of rainfall events’ geometric properties.
Radar-derived rainfall estimation is one of the most significant recent advances in hydrologic engineering and practice. This chapter describes common rainfall interpolation techniques, including Thiessen polygons, inverse distance squared weighting, and kriging. It also shows a rainfall “surface” interpolated from rainfall observed at gauge locations using the Thiessen polygon technique. Distributed hydrologic models are particularly well suited to utilizing radar rainfall. The National Centers for Environmental Information Climate Data Online provides access to NCDC’s archive of global historical weather and climate data in hourly, daily, monthly, seasonal, and yearly time steps. The chapter also presents an overview of the book’s key concepts.
This chapter presents measures used in forecast evaluation procedures common in atmospheric sciences and numerical weather prediction and forecast verification. An exhaustive literature review on bias assessment in radar-based precipitation estimates reveals that several performance metrics or skills scores are common to various studies and regions. The radar data overestimate the serial autocorrelation at several lags relative to gauges. The chapter provides the possible range of values for indexes and skill scores and ideal or perfect values. The ideal values will help in assessing bias-corrected radar values. The chapter also summarizes all available indexes, measures, and scores and their utilities for bias analyses and discusses the utility of these measures and indexes for real-time operations, modeling, and short-term planning. Bias corrections refer to corrections applied to radar-based rainfall data using the rain gauge.
Front matter pages come before the papers or chapters in a published work and include a title page, copyright information, and a table of contents. This publication’s front matter also includes a preface and acknowledgments.
Hydrologic modeling needs continuous rainfall data without gaps. However, rainfall data gaps of different length are often unavoidable due to random and systematic errors. This chapter focuses on spatial and temporal analysis of rainfall and estimation of missing rainfall data using radar-based rainfall estimates. It discusses functional forms for estimating missing rainfall data with linking radar and rain gauge data and geographically weighted optimization methods. The chapter also describes geospatial grid-based transformation of radar data from one spatial coordinate system to another. Deterministic weighting and stochastic interpolation methods have been used for the spatial construction of rainfall fields or for estimating missing rainfall values in space. However, recent studies have reported limitations of spatial interpolation methods. Correlation-weighting techniques and artificial neural network methods are conceptually superior to deterministic approaches compared with the traditional inverse distance weighting method and its variants.
This chapter summarizes using ground- and radar-based methods to measure rainfall. It focuses on evaluating and improving radar-based rainfall measurement by optimizing functional forms of reflectivity-rainfall relationships because radar rainfall estimates are prone to systematic and random errors. Ground-based measurements are the conventional and direct ways of measuring rainfall that are obtained by a network of rain gauges. The chapter discusses the use of Z-R relationships to estimate rainfall and commonly used Z-R relationships. Incorrectly specified Z-R relationships are less of a problem for non-real-time applications than for real-time data. Analysis of uncertainties in Z-R relationships for different storm events and selection of optimal exponents and coefficients requires selecting several rain gauges within a region. The chapter also presents some issues to consider when optimal or revised Z-R relationships are developed and used to estimate radar-based rainfall.
Common formats for gridded radar rainfall data include ESRI Shapefile, Arc/Info ASCII Grid, Gridded NetCDF, and others, and commercially produced gridded rainfall data sets are available in various temporal and spatial resolutions. The data are typically available in geographic information system-compatible formats that are easily processed for input to hydrologic models. Of special note are two standardized grid systems in common use by U.S. government agencies: The Hydrologic Rainfall Analysis Project grid used by the U.S. National Weather Service and the Standard Hydrologic Grid used by the U.S. Army Corps of Engineers. Radar rainfall data have various purposes. Some uses include real-time hydrologic monitoring, flood forecasting, flood hazard and disaster management, probable maximum precipitation estimation, extreme storm analysis, and many others. Accurate estimation of volumetric watershed outflow is a key objective in hydrologic analysis.
'/%3e%3cpath fill='%23fff' d='m8.751-17.959 10.11 11.373L3.997-9.844l13.94-6.1-7.692 13.129z'/%3e%3c/svg%3e)
