What Rainfall Return Frequency? Investigators T. V. Hromadka II1 , M. Phillips1 , P. Rao2 , B. Espinosa3 , R. Perez3 , M. Barton3 1 Department of Mathematical Sciences, United States Military Academy, West Point, NY 10996, United States 2 Civil and Environmental Engineering, California State University, Fullerton, CA 3 Hromadka & Associates, Rancho Santa Margarita, CA 92688, USA

Overview Six published statistical rainfall studies by federal and state agencies, for a study area in California, are reviewed. The reasons for the differences in their rainfall return frequencies are identified and explained.

Research Goals To review the published hydrometeorological characteristics of rainfall and their statistical trends in a particular localized study area of Los Angeles County, California.

Study area Study area (circled) is in Southern California Topographic high regions and canyon locations in the study area

Research Goals Because of the rainfall trough, statistical estimates of return frequencies of rainfall in the study area are complex, and consequently, the considered three agency estimations of rainfall trends differ in the study area. In this paper, these various analyses are compared and difference is assessed,

Published studies that were reviewed National Weather Service (NWS) National Oceanographic and Atmospheric Administration (NOAA): Atlas 2 (1973) and Atlas 14 (2011). California Department of Water Resources (DWR) 2003 and 2010 reports. Los Angeles County (LAC) Hydrology Manual (1993, 2006).

NOAA Precipitation frequency – Atlas 14 They provide rainfall return frequency estimates for various peak durations throughout large regions of the southwest United States, including the study area. The statistical analysis is based upon the Generalized Extreme Value (GEV) distribution, where statistical parameters are estimated regionally using rainfall gages located within the region.

Rain gages used in NOAA analysis Rain gages used in Atlas 14 (Interior and exterior of study area) Atlas 14 Gages in the study area

Rain gages used in California DWR analysis Rain gages used in DWR analysis (Interior and exterior of study area) DWR Gages in the study area

Rain gages used in Los Angeles County (LAC) analysis Rain gages used in LAC analysis (Interior and exterior of study area) LAC Gages in the study area

Comparison of isohyetals for 50-year 24-hour peak duration LAC Hydrology Manual NOAA Atlas 2

Comparison of 50-year 24-hour rainfalls for rain gages located within the study area (and common to all sources of return frequency estimates)

Comparison of 50-year 24-hour rainfalls for between NOAA, LAC, and CA DWR

Conclusions Because of the significant orographic effects and variations in topography within the study area, significant variations in rainfall quantities are observed, resulting in a challenging situation in the analysis and estimation of peak duration rainfall quantities and their respective return frequency estimates. The LAC study had far fewer gages in the study area. As a result, not only is the LAC analysis based upon a much smaller sample size than the DWR or NOAA analyses, but the detail provided in estimating rainfall return frequency values is less spatially defined by the smaller data set used in the LAC analysis.

Comparison of Radar Data Versus Rainfall data Investigators B. Espinosaa, T.V. Hromadka IIb , R. Pereza a Hromadka & Associates, Rancho Santa Margarita, CA 92688, USA b Department of Mathematical Sciences, United States Military Academy, West Point, NY 10996, United States

Overview Doppler radar rainfall data is an important variable in many hydrologic models This study systematically analyzed the accuracy of three Doppler radar rainfall estimates.

Research Goals For two separate storm events (8/25/2014 and 9/8/2014), the published rainfall values from three Doppler radars are compared with the gage data, for an area in Southern California. Possible errors in radar estimates due topographic interference and the need for ‘ground-truthing’ of radar data is described.

Study Location and Rain gage The area is in City of La Quinta, CA. The area is a desert environment municipality at the foot of Santa Rise Mountains on the floor of the Coachella Valley. The rain gage in the study area (shown in the figure, Gage 296) is an ALERT gage and produces a continuous rainfall record.

Three Doppler radars in the study area

Analysis of rainfall data (storm date August 25, 2013) Radar: KSOX (Santa Ana) Rainfall at Gage = 0.1 in Radar: KNKX (San Diego) Rainfall at Gage = 1.17 in Radar: KYUX (Yuma) Rainfall at Gage = 2.34 in

Analysis of data (storm date August 25, 2013) Rain Gage 296 KSOX Santa Ana KNKX San Diego KYUX Yume Storm Total (inch) 2.32 0.10 1.17 2.34 Peak hour (inch) 2.08 1.06 2.09 KYUX Radar rainfall estimate and peak hour measurements correlate well with the gage data

Analysis of data (storm date September 8, 2014) Radar: KSOX (Santa Ana) Rainfall at Gage = 0.04 in Radar: KNKX (San Diego) Rainfall at Gage = 1.18 in Radar: KYUX (Yuma) Rainfall at Gage = 2.19 in

Analysis of data (storm date September 8, 2014) Rain Gage 296 KSOX Santa Ana KNKX San Diego KYUX Yume Storm Total (inch) 3.08 0.04 1.28 2.19 Peak hour (inch) 2.84 1.26 2.09 None of the three radar rainfall and peak hour values correlate well with the gage data (KYUX data)

Radar stations and gage topography Topographic interference (echo effect) of the radar waves from KNKX and KSOX radars led to highly inaccurate radar rainfall estimates.

Conclusions Topographic interference can lead to incorrect radar rainfall estimates and ground-truthing of radar estimates may be required for many storms in certain geographical areas.

Thank You