Extreme Quantitative Precipitation Forecast Performance at the Weather Prediction Center from 2001 to 2011

Abstract

Extreme quantitative precipitation forecast (QPF) performance is baselined and analyzed by NOAA's Hydrometeorology Testbed (HMT) using 11 yr of 32-km gridded QPFs from NCEP's Weather Prediction Center (WPC). The analysis uses regional extreme precipitation thresholds, quantitatively defined as the 99th and 99.9th percentile precipitation values of all wet-site days from 2001 to 2011 for each River Forecast Center (RFC) region, to evaluate QPF performance at multiple lead times. Five verification metrics are used: probability of detection (POD), false alarm ratio (FAR), critical success index (CSI), frequency bias, and conditional mean absolute error (MAE cond ). Results indicate that extreme QPFs have incrementally improved in forecast accuracy over the 11-yr period. Seasonal extreme QPFs show the highest skill during winter and the lowest skill during summer, although an increase in QPF skill is observed during September, most likely due to landfalling tropical systems. Seasonal extreme QPF skill decreases with increased lead time. Extreme QPF skill is higher over the western and northeastern RFCs and is lower over the central and southeastern RFC regions, likely due to the preponderance of convective events in the central and southeastern regions. This study extends the NOAA HMT study of regional extreme QPF performance in the western United States to include the contiguous United States and applies the regional assessment recommended therein. The method and framework applied here are readily applied to any gridded QPF dataset to define and verify extreme precipitation events. * Current affiliation:

Climatology and Environmental Characteristics of Extreme Precipitation Events in the Southeastern United States

“…Several prior studies (e.g., Johnson 2005, 2006;Kunkel et al 2012) have identified EPEs in the United States using the historical gauge-based recurrence interval precipitation thresholds calculated by Hershfield (1961), while others (e.g., Brooks and Stensrud 2000;Ralph and Dettinger 2012;Hitchens et al 2012Hitchens et al , 2013 have used fixed precipitation thresholds. We opted to use geographically varying upper quantiles of daily (24-h period ending 1200 UTC) precipitation amount, similar to Ralph et al (2010) and Sukovich et al (2014). Specifically, the 99th and 99.9th percentile values computed at each grid point for all days in all seasons during 2002-11 with .0 mm of precipitation ( Fig.…”

Large-scale patterns of daily precipitation extremes on the Iberian Peninsula

“…Predicting extreme precipitation events is very important to minimize risks and to plan and establish appropriate management and early warning systems. Various studies have shown that most weather forecast models still poorly predict heavy rainfall events (Fritsch and Carbone, ; Sukovich et al , ), especially those at small spatial and temporal scales. Moreover, model projections of heavy precipitation extremes include substantial uncertainties (Fischer et al , ).…”

“…Moreover, model projections of heavy precipitation extremes include substantial uncertainties (Fischer et al, 2014). Sukovich et al (2014) found less forecasting skill for extreme events than for lower precipitation thresholds in some models, and the skill tended to decrease with longer lead time. One possible alternative approach to this problem is the examination of synoptic patterns, which plays a major role in heavy rainfall events and can therefore be used objectively to help weather forecasters issue earlier warnings for such events (Dolif and Nobre, 2012).…”

A hydrological prediction system based on the SVS land-surface scheme: efficient calibration of GEM-Hydro for streamflow simulation over the Lake Ontario basin

“…As a consequence, no effort was made to represent in a detailed manner the artificial structures of the region in WATROUTE. Moreover, the small diversions occurring to fill some canals in the region, or even the aquifers which can contribute significantly to baseflow (Singer et al, 2003;Kassenaar and Wexler, 2006), do not prevent lumped models from registering good performance, which is helpful to this study.…”

“…Given the continuous increase in precipitation forecast skill of numerical weather prediction (NWP) systems (Sukovich et al, 2014), it became possible to obtain skillful runoff forecasts directly from NWP model outputs, and streamflow forecasts by routing these gridded runoff fields. Indeed, modern NWP models tend to simulate to some extent the snow, vegetation, and soil processes that contribute to the generation of runoff and streamflow.…”