Assessing Two Approaches for Enhancing the Range of Simulated Scales in the E3SMv1 and the Impact on the Character of Hourly US Precipitation. Issue 4 (18th February 2022)

• Record Type: Journal Article
• Authors: Kooperman, G. J.; Akinsanola, A. A.; Hannah, W. M.; Pendergrass, A. G.; Reed, K. A.
• Is Part Of: Geophysical research letters. Volume 49:Issue 4(2022)
• Page Start: n/a
• Page End: n/a
• Publication Date: 2022-02-18
• Subjects: precipitation -- earth system model -- energy exascale earth system model -- high resolution -- multiscale modelling framework -- United States; Geophysics -- Periodicals; Planets -- Periodicals; Lunar geology -- Periodicals; 550
• DOI: 10.1029/2021GL096717
• Languages: English
• ISSNs: 0094-8276

Abstract

Abstract: Improving the representation of precipitation in Earth system models is essential for understanding and projecting water cycle changes across scales. Progress has been hampered by persistent deficiencies in representing precipitation frequency, intensity, and timing in current models. Here, we analyze simulated US precipitation in the low‐resolution (LR) configuration of the Energy Exascale Earth System Model (E3SMv1) and assess the effect of two approaches to enhance the range of explicitly resolved scales: high‐resolution (HR) and multiscale modeling framework (MMF), which incur similar computational expense. Both E3SMv1‐MMF and E3SMv1‐HR capture more intense and less frequent precipitation on hourly and daily timescales relative to E3SMv1‐LR. E3SMv1‐HR improves the intensity over the Eastern and Northwestern US during winter, while E3SMv1‐MMF improves the intensity over the Eastern US and summer diurnal timing over the Central US. These results indicate that both methods may be needed to improve simulations of different storm types, seasons, and regions. Plain Language Summary: Extreme storms and precipitation are expected to become more intense with climate change. However, current global‐scale numerical models that are used for climate projections often misrepresent the intensity and timing of precipitation compared to observations, which can lower confidence in projected changes. The sources of these deficiencies are associated with the low‐resolution and simplified representation of physical processes (e.g., convection) used in these models in order to make century‐long simulations computationally feasible. Here, we investigate how enhancing the range of scales represented in the Energy Exascale Earth System Model (E3SMv1) can improve precipitation. We focus on two configurations that incur significant but similar computational cost and have drastically different approaches to enhance the range of explicitly represented scales: high horizontal‐resolution (HR; ∼25‐km) and multiscale modeling framework (MMF; 2‐km cloud‐resolving models embedded within each grid column of E3SMv1). Both methods improve the frequency and intensity of precipitation over the United States. However, due to the different scales represented by each method, improvements occur in different seasons and regions, primarily during winter in E3SMv1‐HR and summer in E3SMv1‐MMF. These results indicate which configuration may be most useful for studies of different storm types and suggest both methods may be needed to represent precipitation overall. Key Points: Multiscale modeling framework (MMF) and high‐resolution (HR) capture more extreme and less frequent precipitation than conventional E3SMv1 E3SMv1‐HR improves intensity and timing of precipitation in the Eastern and Northwestern US during winter in association with ETCs and ARs E3SMv1‐MMF improves intensity of precipitation in the Eastern US and summer diurnal timing in the Central US related to propagating MCSs