Climate Change, Atmospheric Rivers, and Future

Michael D. Dettinger , Ph.D. Tel: (858) 822-1507 Research Hydrologist Fax: (858) 822-2028 U.S. Geological Survey Email: [email protected] Scripps Institution of Oceanography Web: http://tenaya.ucsd.edu/~dettinge La Jolla, CA 92093-0224

BIOGRAPHICAL SKETCH

Dr. Michael Dettinger is a research hydrologist for the U.S. Geological Survey, Branch of Western Regional Research, and a research associate of the Climate, Atmospheric Sciences and Physical Oceanography Division at Scripps Institution of Oceanography. He has degrees from the University of California, San Diego (Physics), Massachusetts Institute of Technology (Civil Engineering), and a Ph.D. from the University of California, Los Angeles (Atmospheric Sciences). Dettinger has monitored, evaluated, and researched the water resources of the West for over 25 years, focusing on regional surface water and groundwater resources, watershed modeling, causes of hydroclimatic variability, and climatic-change influences on western water resources. Among other activities, he was State Ground-Water Specialist for the USGS in the 1980s, received a Vice President’s National Performance Review Award for physical-sciences leadership in Mojave Desert Ecosystems science and data management planning efforts in 1996, was an Associate Editor of the journal Water Resources Research from 1998 to 2000, was the program chair and fundraiser for the annual Pacific Climate (PACLIM) Workshops, 1998-2004, is a founding member of the multi-institution CIRMONT Western Mountain Climate Sciences Consortium, and served on the CALFED Water Management Science Board.

ABSTRACT

Historically, the most dangerous storms in California have been warm wet storms that strike in winter, producing intense over large areas and unleashing many of the State’s largest floods. The most commonly recognized of these storms have been described as “pineapple express” storms because of the way that they appear (in weather satellite imagery) to draw warm, moist air from the tropics near Hawaii northeastward into California. Recent studies, though, have shown that pineapple express storms are just one version of a common feature of midlatitude weather, called “atmospheric rivers” (ARs). We now know that, globally, about 90% of all the water vapor transported towards the poles across the midlatitudes is transported within the narrow, intense filamentary bands of moist air that form these ARs. Because AR storms are increasingly understood to have been the source of most of the largest floods in California, an evaluation of the future of floods under climate change must attempt to project the future frequencies and intensities of ARs.

Using a locally-based strategy for detecting AR-type storms along the California coast, developed at the NOAA Earth System Research Lab, climate simulations from seven global-climate models (GCMs) were analyzed to compare frequencies and magnitudes of AR storms arriving in California under simulated historical and climate-changed conditions. First, numbers of AR episodes in the climate models and in the observational record were compared to find that, although on average most of the models generate more ARs than observed, the general distribution of AR days per winter were not so different as to preclude evaluations of the projected changes. Next, in comparing historical to future climate simulations, changes in AR storms in the models were found to occur mostly at the extremes: Years with many AR storms become more frequent in most of the climate-change projections, but the average number of such storms per year are not projected to change much. Similarly, although the average intensity of the storms is not projected to increase much in most models, occasional much-stronger-than-historical-range storm intensities are projected to occur under the warming scenarios. The simulated AR storms also warm along with the winter- mean temperatures in the seven models. Together these findings suggest that California risks from the warm-wet, atmospheric-river storms may increase beyond those that we have known historically, mostly in the form of occasional more-extreme-than-historical storm seasons. CLIMATE CHANGE, ATMOSPHERIC RIVERS & FUTURE CALIFORNIA FLOODS

Michael Dettinger (USGS) with contributions from Marty Ralph (NOAA) & Dale Cox (USGS)

Extreme Precipitation Symposium Sacramento, CA 24 June 2009 Outline

• Perspectives on Atmospheric Rivers (ARs) • Detecting Atmospheric Rivers • ARs in climate-change projections • Conclusions Atmospheric Rivers as Hazards

OF MORE THAN 40 CM Relationship of ARs and Flooding

Atmospheric river

From Ralph, et al., 2006

Dettinger, Extr Precip Sympos, 2004 Atmospheric Rivers as Water Supply

Atmospheric river

From Ralph, et al., 2006 Atmospheric Rivers as Water Supply

California AR Landfalls

Atmospheric river

From Ralph, et al., 2006 Atmospheric Rivers as Water Supply

Atmospheric river

From Ralph, et al., 2006 Recognizing Atmospheric Rivers Atmospheric river

Vertically Integrated Water Vapor Transports

SSMI Vertically IVT > 500 kg/m/s Integrated Water Vapor Content (> 2000 km long; < 1000 km wide; IWV > 2 cm) Recognizing Atmospheric Rivers Atmospheric river Observed Rainfall Intensity (colors) vs Forcings (upslope wind) and Fuel (water vapor) at Cazadero

Rain >10 mm/h: >12.5 m/s; >2 cm

From Ralph, et al., 2006

Neiman et al 2009 Recognizing Atmospheric Rivers Atmospheric river Simulated Forcings (upslope wind) and Fuel (water vapor) on Central CA Coast Future Atmospheric Rivers Atmospheric river

Cayan et al, 2009, CEC Future Atmospheric Rivers Atmospheric river Simulated Forcings (upslope wind) and Fuel (water vapor) on Central CA Coast Future Atmospheric Rivers

Numbers of AR storms on Central California coast in 7 climate-change projections & historical record

From Ralph, et al., 2006 Future Atmospheric Rivers

Numbers of AR storms on Central California coast in 7 climate-change projections & historical record Average # yrs < # yrs > # yrs > # per yr 5 ARs 15 ARs 20 ARs Reanalysis 5.8 days/yr 42 % of yrs 3 % of yrs 0 % of yrs 1961-2000 Projections 1961-1980 8.5 25 16 5 1981-2000 9.0 27 16 8 2046-2065 11.6 12 28 10 2081-2100 11.7 16 32 12 30% 53% 100% 85% increase decrease increase increase Future Atmospheric Rivers

Projected trends in numbers of AR storms on Central California coast in 7 climate-change projections

CCC CNRM ECHAM GFDL GISS MIRO MRI (Canada) (France) (Germany) (USA) (USA) C (Japan (Japan ) ) Change in + 7 +2 +4.5 +0.4 +0.3 +2 +4 number of days days ARs/yr in days days days days days 21st Century Future Atmospheric Rivers

Expect more moisture content in many storms Future Atmospheric Rivers

Expect weaker upslope winds in many storms Future Atmospheric Rivers

Let Product of Vapor * Winds = Intensity

Projected: Not so much change in mean intensities, but more extreme outliers Future Atmospheric Rivers

Broad warming of AR storms, with some warmer than any historicals Future Atmospheric Rivers

From Ralph, et al., 2006 Projected: AR warming along with general winter temperatures, but not as fast Future Atmospheric Rivers

Lengthening of AR seasons?

Observed (Dettinger 2004) Summary

• The future of landfalling atmospheric rivers is important for the future of flood hazards AND water resources in California Projections of 21st Century climates suggest: • More years with lots of ARs, fewer with few • Moister ARs with weaker upslope winds • Overall average intensities don’t change much but occasional much stronger than historical ARs • ARs warmer by about +2C on ensemble avg • AR season extends Implication

• Thus, preparations for extreme storms (like the one being dramatized by the ARkStorm effort) is all the more prudent in view of current climate- change projections

For broader discussion of projected flood regimes, see:

Dettinger, M.D., Hidalgo, H., Das, T., Cayan, D., and Knowles, N., 2009, Projections of potential flood regime changes in California: California Energy Commission Report CEC-500-2009-050-D, 68 p.