72 Expected Future Precipitation Changes from the North American Monsoon

The Great Salt Lake Basin Integrated Plan (GSLBIP) is used to determine water allocation across the greater Salt Lake region. Development of a tradeoff analysis for the basin plan is contingent upon a greater understanding of water entering the system through precipitation. Global climate models (GCMs) are comprised of assumptions based on an initial condition and indication of change in future greenhouse gas emissions. They return predicted future variables, such as precipitation. However, different GCMs vary greatly in expected precipitation for June, July, and August (JJA) at the end of the century. Particularly wet models for a high greenhouse gas emission scenario [Shared Socioeconomic Pathway (SSP) 5.85; (O’Neill et al., 2014)] predict up to a doubling of precipitation in the summer from the historical period, 1979-2014, to the future period, 2070-2099, while the driest model for SSP 5.85 simulates precipitation being just above half of historical values. The Clausius Clapeyron relation (Martinkova et al., 2020) between temperature and atmospheric humidity indicates that as the climate system warms, the capacity for moisture in the atmosphere increases. Increased evaporation also occurs with rising temperatures. Climate is predicted to warm by the end of the century, however the impact of this on summer precipitation has not been studied in detail for the Great Basin area. The large range in projected JJA precipitation is likely due to contrasts among GCMs in simulations of changes in the North American Monsoon (NAM).

Since the GSLBIP focuses on a specific region, high resolution data must be used to analyze precipitation values across the study region in enough detail to be useful. Our team downscaled GCM data for the Coupled Model Intercomparison Project Phase 6 (CMIP6) (Eyring et al., 2016) from 1° (~111-km) to 4-km grid spacing following the Multivariate Adaptive Constructed Analogs (MACA) method (Abatzoglou et al., 2012) due to its ability to represent near surface meteorological patterns in complex orographic terrain (see Figure 1 for difference between GCM and MACA). MACA data were analyzed over the study region and confirmed to reflect the same climate signals for precipitation change found in GCM data (Figure 2). Since the MACA and GCM data both capture a divergence in summer mean precipitation ratios, we returned to GCM data to better understand what greater climate patterns might be causing this.

We chose to focus our analysis on models that produced extreme precipitation ratios. Wet models consisted of the four models with the highest precipitation ratios for SSP 5.85 (UKESM1-0-LL, ACCESS-CM2, CanESM5, KACE-1-0-G). Dry models consisted of the four lowest (MPI-ESM1-2-LR, CNRM-ESM2-1, CNRM-CM6-1-HR, INM-CM4-8). Data from wet and dry models were individually mapped as means across the future and historical period and as anomalies. Mapped areas extended over the continental US as well as parts of the Pacific Ocean using GCM data to capture climatological impacts of the NAM on the study region. The variables, precipitation, geopotential height, sea level pressure, and wind vectors, were analyzed in their groups for common patterns. Each variable was specifically chosen because of its ability to interpret how the NAM was modeled.

Two clear and logical patterns appeared across the wet models that explain a future increase in precipitation. (1) The concentrated area of low pressure over the land mass shifts northward while the subtropical high pressure area over the Pacific Ocean is reduced causing a weakening of the climatological southward winds (Figure 3a). This allows for an increase of moisture advection into the western US. Pressure is one of the main drivers of climate patterns in the atmosphere. Monsoons are dictated by pressure as air is accelerated from areas of high to low pressure to reach an equilibrium. However, pressure is also greatly influenced by temperature.

Areas of high surface temperature tend to have low pressure because air touching the surface is heated and begins to rise. This causes a low pressure system over land due to its relatively low specific heat. Over the Pacific Ocean, we can see an area of high pressure in dark red referred to as the subtropical high (Figure 3b). Notice that wind vectors move in a clockwise motion around high pressure regions and counterclockwise around low pressure regions due to the Coriolis effect in the Northern Hemisphere. Strong southward winds from these patterns move along the west coast inhibiting moisture from traveling further inland. Changes in pressure (Figure 3a) reduced the climatological southward flow, potentially making it easier for storms to move further inland.

(2) Amplified ridging, reflected as increased geopotential heights, over the central plains (Figure 3c) may create conditions conducive for stalling summer storms over the study area. Geopotential height is used to measure how high you must go into the atmosphere to reach a certain pressure. As temperature increases, a column of air expands reducing its pressure and increasing its geopotential height. Areas of low geopotential height represent troughs while areas of high geopotential height represent ridges. Because of the climatological westerly winds across the continental US, weather systems tend to move from west to east. Downstream from troughs tend to be areas of surface low pressure with heavy precipitation, particularly if ridging causes a barrier for eastward progress of the storm. Heating over land is projected to cause increased ridging over the continent in the future simulations. The area of highest geopotential height (shown in dark red, Figure 3c) has shifted northward as well as the angle of the western flank steepening (i.e., the contours become more north-south in orientation, reflecting a strengthened west-east gradient). Meridional ridges and troughs (formations amplified in the north and south along meridians) are more apt at trapping storms over the western US. Further research needs to be conducted to conclude if increased precipitation in the area is due to longer-duration storm events from ridging, an increased frequency of storms moving into the area, or increased availability of precipitable water.

In summary for the wettest models, the combination of altered pressure gradients and increased atmospheric ridging creates a more favorable environment for moisture-rich storms to penetrate further inland and remain over the study area. This is reflected in the summer precipitation anomalies (Figure 4), which align with both enhanced Pacific moisture inflow and increased rainfall associated with stalled midlatitude cyclones over the Great Salt Lake Basin.

While these clear and consistent patterns appear in the wet models, we need further study to understand the climate patterns underlying variations among the dry models. We also plan to further investigate extremes of daily precipitation in the MACA data, focusing on metrics relevant to stakeholders, including indicators of drought (e.g., consecutive dry days) and extreme convective precipitation.

Bibliography

Figures