By Melanie Lenart | The University of Arizona | September 14, 2008
Although climate change projections are most reliable at the global level, people seeking to adapt and respond to the projected changes need to know about climate change impacts at the local level. Fortunately, increasing resolution and other advances in Global Climate Models (GCMs) are helping to improve regional scale modeling (Figure 1).
Figure 1. Average annual precipitation patterns in Arizona (a) tend to reflect elevation of the landscape, with peaks occurring along mountain ranges. A regional model of Arizona precipitation with a scale of about 350 square miles per grid square (b) would average out some of the precision, but retain much more information than found in a global climate model.
| Enlarge This Figure |
Credit: PRISM Group, Oregon Climate Service, Oregon State University
A general rule holds that a regional model’s resolution should not surpass about one-twelfth of the resolution of the model feeding into it, typically a GCM.1 As GCM resolution has increased to about 4,000 square miles per grid square, regional models could reliably drop down to resolutions of about 350 square miles, roughly twice the area of Albuquerque. Scientists are confident resolution will continue to improve as computational power increases.
With better spatial resolution, regional scale models can help resolve important factors affecting Southwest climate, including:
- Jet stream activity, which particularly influences the frontal storms driving winter and spring precipitation
- Tropical storms comprising the monsoon and remnant hurricanes in summer and fall
- Elevation effects, with their year-round influence on temperature and precipitation
Jet streams high in the atmosphere direct where moisture lands. The polar jet stream occurs consistently along the polar front, where cold air meets warm air, fueling many mid-latitude storms. The jet stream occasionally dips down into the subtropics, influencing precipitation in the Southwest.
While researchers have confidence in the models’ ability to project changes in the polar jet stream and storms in the mid-latitudes (roughly 40 to 70 degrees north of the equator), they recognize some deficiencies exist. Potential problems include inadequacies where the mid-latitudes interact with the tropics—namely, the subtropics.1
In addition, modelers have found it challenging to capture the historic year-to-year variability of the northern jet streams.1 Several research teams have shown a poleward shift of the northern jet stream, linked to the expansion of Hadley cell circulation.2,3 Researchers at The University of Arizona have also linked the poleward shift of the jet stream to drier springs in the time frame they considered, from 1978 to 1998.4
The jet stream is more likely to dip south into the Southwest during El Niño years, wielding the largest effect on winter precipitation. Thus, recent improvements in modeling El Niño variability in the distant tropics have implications for modeling regional jet stream activity and its effect on winter precipitation.
Still, only about a third of the 18 GCMs tested were identified as having a relatively realistic interpretation of El Niño fluctuations.1 The most skillful models tend to have relatively high resolution—with some as fine as 500 square miles—in the tropical Pacific, where El Niño has its greatest effect on sea surface temperature.1
In addition, the poleward shift of the jet stream observed for the past several decades means that the regional impacts of El Niño could change along with global climate. The spring drying identified by University of Arizona researchers could be related to this poleward shift in the jet stream related to global warming.4
Improvements in resolution, which is highest in regional and local models, can also lead to better depictions of convective storms. These include the storms that comprise the monsoon and hurricanes.
As it happens, the models that do a better job at reproducing El Niño conditions also tend to show more realistic variability of the Asian monsoon.1 This relates in part to the interactions between these two climate patterns. In Asia as in the Southwest, conditions that favor El Niño development tended to weaken the monsoon, although these interactions can change with climate. They also affect other tropical storms, including hurricanes.
In general, GCMs tend to underestimate the magnitude of precipitation events.5 In other words, they’re not as good at capturing extremes, such as droughts and floods, as they are at modeling average conditions.
Some of this relates to resolution. At the scale of a GCM grid cell, typically 4,000 square miles or larger in current models, rainfall registers frequently and moderately (Figure 1).
That’s partly because precipitation events at that scale tend to be spotty, especially in the tropics and subtropics. So flood-producing storms can balance out drought-producing dry spells when occurring in one grid square, giving the misleading appearance of average conditions at the larger scale.
Curious about modeling and the North American Monsoon? Read Monsoon Modeling by Chelsey Killebrew.
Regional models using a resolution of about 350 square miles or below are more likely to capture convective storms, which typically occur on scales even smaller than the typical regional model grid square. Improvements continue as resolution increases, as long as the input is relevant for the smaller scale modeled.
The higher resolution of regional models also improves the land-sea interactions that drive the regional monsoon. Higher-resolution models are more likely to simulate the wind shift that occurs as seasonal heating pushes land temperature above local sea surface temperatures. Higher resolution also improves a model’s potential at capturing changes in local seas, such as the Gulf of California, which blends into the Pacific Ocean under the lower resolution of many GCMs.
Extreme precipitation events are likely to remain difficult to model with regional models as well as GCMs because challenges exist for reasons other than resolution. By definition, extreme events are rare, so assessing how well models capture them remains challenging—especially in an arid climate. For instance, while rainfall exceeding four inches a day might occur roughly once a year in Miami, it might happen only once every two centuries in Phoenix.1
Hurricanes and other tropical cyclones remain poorly resolved in most models, although researchers are optimistic about pending improvements. Some GCMs have been used to project climate-related changes in hurricanes, with more success in simulating frequency than intensity.6 Uncertainties remain regarding changes in frequency, intensity, and storm tracks.
Again, resolution improvements can help but they won’t solve all the problems. For instance, even meteorologists who predict short-term hurricane behavior with highly detailed weather models find it challenging to pinpoint storm path. Sudden shifts in weather patterns can occur that send hurricanes off in a different direction than expected.
The Mogollon Rim runs through Arizona and represents the southern edge of the Colorado Plateau. The Rim is a significant landscape feature affecting regional weather patterns and is important to resolve for improving climate change projections.
Credit: ©Jim Wark, Airphoto
With their higher resolution, regional models could improve the depiction of local climate patterns affected by elevation. This could even help resolve mountain climates, depending on how landscape features fit into the model’s grid square.
Still, models provide only one average elevation for an entire grid square. Even the current resolution of about 350 square miles, almost twice the size of Tucson, precludes capturing the details of mountain slopes and valleys.
If the GCM or regional model accurately depicts climate patterns affecting an area, though, it might serve to give researchers the information they need to downscale results to the local level of a mountainside.
Research efforts in this area have been successful in capturing fine-scale details of historic weather and climate, suggesting that these methods can add value for assessments of the impacts of climate change projections.7
"Drier, warmer springs in U.S. Southwest stem from human-caused changes in winds", August 19, 2008, UA News
Human-driven changes in the westerly winds are bringing hotter and drier springs to the American Southwest, according to new research from The University of Arizona. The finding is the first to link the poleward movement of the westerly winds to the changes observed in the West's winter storm pattern.
| http://uanews.org/node/21017 |
- Bader, D.C., et al. 2008. Climate Models: An Assessment of Strengths and Limitations. A Report by the U.S. Climate Change Science Program and the Subcommittee on Global Change Research. Department of Energy, Office of Biological and Environmental Research, Washington, D.C., 124.
- Seidel, D.J. and W.J. Randel. 2007. Recent widening of the tropical belt: Evidence from tropopause observations. Journal of Geophysical Research, 112: D20113, 1-6.
- Hu, Y. and Q. Fu. 2007. Observed poleward expansion of the Hadley circulation since 1979. Atmospheric Chemistry and Physics, 7: 5229–5236.
- McAfee, S.A., and J.L. Russell. 2008. Northern Annular Mode impact on spring climate in the western United States. Geophysical Research Letters,. 35 (L17701): 10.1029/2008GL034828.
- Allan, R.P., and B.J. Soden. 2008. Atmospheric warming and the amplification of precipitation extremes. Science, 321: 1481–1484.
- Christensen, J.H., et al. 2007. Regional climate change projections. In: Climate Change 2007: The Physical Science Basis. Contribution of Working Group I to the Fourth Assessment Report of the Intergovernmental Panel on Climate Change. [Solomon, S., D. Qin, M. Manning, Z. Chen, M. Marquis, K.B. Averyt, M. Tignor, and H.L. Miller (eds.)] Cambridge University Press, Cambridge, United Kingdom and New York, NY, USA.
- Maurer, E.P. and H.G. Hidalgo. 2008. Utility of daily vs. monthly large-scale climate data: an intercomparison of two statistical downscaling methods. Hydrology and Earth Systems Sciences Discussions, 4: 3413–3440.