Skill of Rain–Snow Level Forecasts for Landfalling Atmospheric Rivers: A Multimodel Assessment Using California’s Network of Vertically Profiling Radars

Abstract

The partitioning of rain and snow during atmospheric river (AR) storms is a critical factor in flood forecasting, water resources planning, and reservoir operations. Forecasts of atmospheric rain–snow levels from December 2016 to March 2017, a period of active AR landfalls, are evaluated using 19 profiling radars in California. Three forecast model products are assessed: a global forecast model downscaled to 3-km grid spacing, 4-km river forecast center operational forecasts, and 50-km global ensemble reforecasts. Model forecasts of the rain–snow level are compared with observations of rain–snow melting-level brightband heights. Models produce mean bias magnitudes of less than 200 m across a range of forecast lead times. Error magnitudes increase with lead time and are similar between models, averaging 342 m for lead times of 24 h or less and growing to 700–800 m for lead times of greater than 144 h. Observed extremes in the rain–snow level are underestimated, particularly for warmer events, and the magnitude of errors increases with rain–snow level. Storms with high rain–snow levels are correlated with larger observed precipitation rates in Sierra Nevada watersheds. Flood risk increases with rain–snow levels, not only because a greater fraction of the watershed receives rain, but also because warmer storms carry greater water vapor and thus can produce heavier precipitation. The uncertainty of flood forecasts grows nonlinearly with the rain–snow level for these reasons as well. High rain–snow level ARs are a major flood hazard in California and are projected to be more prevalent with climate warming.

Freezing Level Forecast Error Can Consume Reservoir Flood Control Storage: Potentials for Lake Oroville and New Bullards Bar Reservoirs in California

“…This amplified difference owes to the differences in the watershed drainage area and in the associated reservoir flood pool storage. This outcome highlights the importance of accounting for the watershed's drainage area and reservoir capacity in estimating the watershed runoff and hence the reservoir storage responses, in addition to the precipitation total, duration, and intensity (Lamjiri et al, 2017; Ralph et al, 2019) as well as the Z FL and its forecast error (Henn et al, 2020).…”

“…Henn et al (2020) found average Z FL forecast errors of ~300–350 m for 30‐ to 72‐hr lead time over the Northern Sierra Nevada during 2017. Based on this finding, we apply a Z FL error of ±350 m to both watersheds and their reservoirs in our analyses (Table 1 and Figure 1a).…”

“…Like all numerical weather prediction information, Z FL forecast is prone to errors and uncertainty. Based on a comparison between 19 profiling radars and the forecasts generated by Version 2 of the Global Ensemble Forecast System (Hamill et al, 2013), the National Weather Service's (NWS) California Nevada River Forecast Center (CNRFC), and the West Weather Research and Forecasting model (West‐WRF) (Martin et al, 2018) during the 2016–2017 winter season, Henn et al (2020) found an average Z FL error of 342 m for lead times of ≤24 hr over Sierra Nevada watersheds. This error generally grew with increasing lead time.…”

“…This error generally grew with increasing lead time. Moreover, Z FL forecast errors increased with heavier precipitation events and higher Z FL , on the top of the fact that Z FL forecasts were low biased for the warmest and wettest events (Henn et al, 2020; Neiman et al, 2014; White et al, 2010). These factors collectively render flood forecasting a greater challenge than estimating precipitation accumulation alone.…”

“…The objective of this study is to describe the key parameters dictating the watershed's runoff response to Z FL forecast error in observed precipitation events, with a goal of informing both hydrologic research and operational stakeholders. Specifically, we investigate the watershed total runoff volume uncertainties as a response to the scale of Z FL error estimated by Henn et al (2020), across a range of observed precipitation distributions in the Feather and North Fork Yuba River watersheds in Northern California. Lastly, we evaluate the associated flood pool storage sensitivities of Lake Oroville and New Bullards Bar reservoirs to precipitation under variable Z FL conditions.…”