Our Skill in Modeling Mountain Rain and Snow is Bypassing the Skill of Our Observational Networks

Abstract

17In mountain terrain, well-configured high-resolution atmospheric models are able to 18 simulate total annual rain and snowfall better than spatial estimates derived from in situ 19 observational networks of precipitation gauges, and significantly better than radar or 20 satellite-derived estimates. This conclusion is primarily based on comparisons with 21 streamflow and snow in basins across the Western United States and in Iceland, Europe, 22and Asia. Even though they outperform gridded datasets based on gauge-networks, 23 atmospheric models still disagree with each other on annual average precipitation and 24 often disagree more on their representation of individual storms. Research to address 25 these difficulties must make use of a wide range of observations (snow, streamflow, 26 ecology, radar, satellite) and bring together scientists from different disciplines and a wide-27 range of communities. 28 29 30 Capsule 31We have now crossed a threshold where, for many mountain ranges, well-configured high-32 resolution atmospheric models are better able to represent range-wide total annual 33 precipitation than the collective network of precipitation gauges, i.e., observations. 34 35 3 Lundquist et al.: Orographic precip: Models vs. Obs 36 Introduction 37We have now crossed a threshold where, for many mountain ranges, well-configured 38 high-resolution atmospheric models are better able to represent range-wide total annual 39 precipitation than the collective network of precipitation gauges, i.e., observations. The 40 prior sentence is disturbing. If we even assign some truth to the statement "Models are 41 better than observations," where does that lead us? If two models are "better than 42 observations" but disagree with each other, which one should we trust more? What do we 43 mean by "better," and what counts as an "observation"? Generalities are dangerous, and for 44 which models, which observations, and which times and locations might this statement be 45 true? How do we identify these specificities in a way that allows us to move forward, 46 scientifically and objectively, in a situation where the truth is difficult to discern? 47Here, we review recent research that collectively suggests, at least for the mid to 48 northern latitudes, that modeled-precipitation has crossed a threshold in range-wide 49 accuracy relative to observation-based-precipitation datasets in complex terrain (Sections 50 2 and 3). We carefully examine the basis for this conclusion, which often consists of 51 multiple indirect observations. We then propose that crossing this threshold requires a 52 fundamental shift in how hydrologists and atmospheric scientists interact. In the past, 53 gridded precipitation datasets that interpolated between existing observations, such as 54 PRISM (Daly et al. 2008), provided a stable medium of both the best available input to a 55 hydrologic model and the best available benchmark for an atmospheric model (Figure 1). 56These gridded datasets are often referred to as "observations," when in truth th...

Changes in the future summer Mediterranean climate: contribution of teleconnections and local factors

“…The climatology in CM2.5, in terms of pattern and magnitude, seems to be more consistent with the EOBS dataset. However, due to a relatively large observational uncertainty in many mountainous areas leading to the underrepresentation of precipitation over complex terrain in gridded observational datasets, it is difficult to validate the model rainfall climatology in the region (Lundquist et al, 2019). The CM2.5 results are comparable to the downscaling simulations using high-resolution (at ∼ 50 and ∼ 12 km) regional climate models of the EURO-CORDEX experiment .…”

“…In the summer, the northward shift of the Hadley cell reveals a connection between the hot and arid eastern part of the Mediterranean and the Asian and African monsoons, as well as a possible connection between these two monsoons (Rodwell and Hoskins, 1996;Ziv et al, 2004;Fontaine et al, 2011;Raicich et al, 2003;Rowell, 2003). The thermal balance of the central-eastern part of the Mediterranean is largely maintained by the two dynamical factors, i.e., the cool air advection of the low-level northerly winds (i.e., Etesians; HMSO, 1962;Metaxas, 1977;Maheras, 1980;Prezerakos, 1984;Reddaway and Bigg, 1996;Zecchetto and de Biasio, 2007;Chronis et al, 2011) and the adiabatic warming of the mid-and upper-level subsidence winds (Raicich et al, 2003;Mariotti et al, 2002;Tyrlis et al, 2013), which counterbalance each other. Ziv et al (2004) have shown these two factors to be significantly correlated, pointing to the Asian summer monsoon, which exerts an influence on the Mediterranean surface, middle, and upper troposphere dynamics.…”

Projected Changes to Extreme Runoff and Precipitation Events From a Downscaled Simulation Over the Western United States

“…Much work has been done to compare WRF-simulated precipitation to both in situ and gridded datasets (Caldwell et al, 2009;Leung and Qian, 2009;Qian et al, 2010;Gutmann et al, 2012;Cardoso et al, 2013;Warrach-Sagi et al, 2013), and the consensus is that WRF estimates are reasonably skilled at capturing realistic precipitation patterns and magnitudes (Hughes et al, 2017). Recent work suggests that atmospheric models such as WRF provide better estimates of precipitation in mountainous regions than precipitation gauge networks (Lundquist et al, 2019); known errors in observational networks pose major challenges to validating climate simulations. Therefore, we do not aim to evaluate WRF-simulated precipitation in this work and instead focus on precipitation changes throughout the 21st century.…”

The Ongoing Need for High-Resolution Regional Climate Models: Process Understanding and Stakeholder Information

“…Improving the representation of orographic forcing also improves simulations of extreme precipitation in mountains, such as that induced by atmospheric rivers (e.g., Leung and Qian 2009;Chen et al 2018). In fact, in mountainous regions, well-configured regional models may produce better estimates of total annual rain and snow than current observational estimates (Lundquist et al 2019) and improve understanding of processes driving surface hydrologic extremes associated with landfalling atmospheric rivers in mountainous areas (Chen et al 2019).…”