Model resolution impact on Precipitation: Comparison to SNOTEL observations

Model resolution impact on Precipitation: Comparison to SNOTEL observations 5/7/2012 40 proj. updates

Sensitivity to model grid resolution: Precipitation from the 8-year Current Climate Simulation Percent model bias in annual total and 8-yr total precipitation [Model Obs] / Obs*100 2000-01 2001-02 2002-03 2003-04 2004-05 2005-06 2006-07 2007-08 8 years 4 km 3.4 15.4-0.4-2.0 5.4 7.3 8.1-2.3 3.9 12 km 28.0 45.1 16.7 20.4 14.5 23.6 26.9 9.2 21.9 36 km 20.2 32.6 4.6 5.9 5.3 8.4 16.7-4.2 10.1 5/7/2012 41 proj. updates

Mean difference in monthly precipitation between WRF and SNOTEL from 8-year climatology data 4 km 12 km 36 km 5/7/2012 42 proj. updates

Model resolution impact on vertical velocity 1 December 2007 0000 UTC Elevation 6 km 36 km

WRF model simulation of Snowpack (Snow Water Equivalent) for two different model resolutions 1 Dec. 2007-1 July 2008 36 km resolution WRF simulation over water year shows complete loss of snowpack by April 1 and the smearing of snowpack across topographic gradients while 2 km simulation shows snowpack to last through the end of July and also produces the correct spatial pattern as compare to the 111 SNOTEL sites (black dots) 36 km 2 km

Headwaters 8-year simulations longer time period for analysis avoid hydrologic spin-up issues Uses Noah 3.2, with the Barlage snow scheme adjustments

Pseudo-Global Warming Simulation to study impact of increased temperature (2 C) and moisture (15%) on the Headwaters water cycle Approach Add mean 10 year average pertubation from current to future climate from NCAR CCSM3 50 year A1B simulation to North American Regional Reanalysis boundary conditions and use to drive the WRF model Rasmussen et al, J. Climate, 2011

Pseudo-Global Warming (PGW) Methodology Schär et al (1996), Sato et al. (2007), Hara et al. (2008), Kawase et al. (2009) 1. Calculate perturbation in 10-yr monthly mean values of U, V, T, geopot. hgt., P sfc and Q v between current and future climate periods from a Climate Global Circulation Model. (SRES-A2 from NCAR CCSM3 CCGM). 2. Add perturbation to current analyses of atmospheric conditions (North American Regional Reanalysis, 3-hrly) and extract regional model initial and lateral boundary conditions Monthly mean of past condition CCSM 1995-2005 Monthly mean of future condition CCSM 2045-2055 Decadal monthly perturbation based on CCSM NARR (initial and 3-hrly boundary conditions) Notes: Sub-monthly phenomenon such as extra-tropical storms not captured except in their mean effect. No change in storm tracks, and thus transient spectra the same (i.e. same climate variability in the future except for

Impact of CO 2 warming on the Water Cycle over the Colorado Headwaters Q = Precip ET or Q = Precip ET (annual average basis, neglecting soil storage) Q Runoff, ET - Evapotranspiration

1 = ET/P +Q/P (annual average basis) Q Runoff, ET Evapotranspiration, P - Precipitation What fraction of P is ET and what fraction is Q? (current and future)

Current Precipitation (average from 8 year simulation) Winter Summer Annual

Future Precipitation (average from 8 year simulation) Winter Summer Annual

Difference in Precipitation: Future Current (average from 8 year run) Winter Summer Annual

Precipitation Difference (future current)

Rain difference (future - current)

Snow difference (future current)

Snow cover difference (future current)

SWE difference (future current)

ET difference (Future Current)

Runoff Difference (Future Current)

Soil moisture difference (Future Current) No net change in annual soil moisture

Seasonal Variations Future precipitation is greater than current in the winter and less than current in the summer. Snowfall decreases in Oct., Nov. April, May and June but increases in December, January, and February. Rainfall increases in the winter and decreases in the summer. Soil moisture dries out in the summer and moistens in the winter, especially above 3000 meters, where snow is resident for most of the year. Evapo-tranpiration is small in the winter but increases rapidly during the spring snowmelt in the future simulation, and is higher than current in the summer at elevations above 3000 meters (soil moisture high at that elevation due to reduced period of evaporation and colder temperatures).

Changes in Seasonality Snowmelt starts two weeks earlier in the future simulation and ends three weeks earlier than current. The early end of snowmelt reduces runoff significantly in June as well as the longer period for evaporation.

Changes in Snowfall, Snowcover and SWE due to climate change Snowfall increases through April 1 due to moister storms. Snowcover decreases by 15% by April 1 due to the enhanced melting of SWE at lower elevations in a warmer climate. SWE similar in the future as the present through February due to the enhances snowfall amount at high elevations compensating for the melting of falling snow and snowpack at lower elevations. SWE decreases by 25% by April 1 due to the early onset of warmer temperatures in a warmer climate.

CO 2 Warming impact on the Water Cycle over the Colorado Headwaters 1 = ET/P +Q/P (annual average basis) Q Runoff, ET Evapotranspiration What fraction of P is ET and what fraction is Q? (current and future) Current: Future:

CO 2 warming impact on the Water Cycle over the Colorado Headwaters 1 = ET/P +Q/P (annual average basis) Q Runoff, ET Evapotranspiration What fraction of P is ET and what fraction is Q? (current and future) Current: 0.81 and 0.19 Future: 0.83 and 0.17

Global Hydrological Cycle Global Land Evap/Precip = 73/113 = 0.65 Runoff/Precip = 40/113 = 0.35 Colorado Headwaters Evap/Precip = 0.81 Runoff/Precip = 0.19 T MES ESSL

Two basic questions for this seminar: 1. Does the predicted increase of precipitation offset the expected increase of evapotranspiration over complex terrain in a warmer climate? 2. How does this impact future runoff? Focus on a snow-melt dominated basin

Current, Future, and Current minus Future Water Balance for Headwaters 530 mm 428 mm 570 mm 473 mm 40 mm 45mm 102 mm 97 mm -5 mm

Impact of Warming on the Annual Water Budget for the Headwaters Q = Precip ET (7 year annual average) ET is increased by 45 mm (+10.5%) Precip is increased by 40 mm (+8 %) Net change in runoff ( Q): - 5mm (-5 %) Both Precipitation and ET increases, but ET increases more leading to a decrease in runoff.

CMIP5 Model-simulated Changes in 21 st Century, RCP4.5 Precipitation Soil Moisture x CMIP5 suggests the Colorado Rockies water cycle will have the following changes: Precipitation: +3% Soil Moisture: -6% Evaporation: +6% Runoff: Runoff -12% WRF-PGW approach suggests the Colorado Rockies water cycle will have the following changes: Precipitation: +8.0% Soil Moisture: 0% Evaporation: +10.5% Evaporation Runoff: -5% What about Thailand? Precipitation:?% Soil Moisture:?% Evaporation:?% Runoff:?% The simulation of orographic convection may be crucial for proper estimation of future precipitation.

Synthesis: How do these results compare to other regional predictions of Colorado River flow? Previous studies (Chrisiensen and Lettemaier 07, Hoerling et. al, Dai et. al, Seager et. al) suggest that the increase in evaporative demand due to warming will outpace potential increases in precipitation subsequently driving increased aridity in the West. Runoff predicted to decrease by 5 to 45%. Current results suggest only small changes in runoff if storm tracks don t change in the future.. Estimation of ET, however is still an issue as well as any changes in storm tracks.. WRF-PGW

Conclusions Both precipitation and evapo-transpiration increase under a warmer, moister climate. The evaporation fraction increases and the runoff efficiency decreases. Note: Summer precipitation decreases, however, yielding a smaller increase in precipitation than anticipated. Without this effect runoff might be positive. Evapo-transpiration increases more than precipitation, yielding a negative change in runoff. Note, however, the uncertainty in the parameterization of evapo-transpiration is large and can easily change the sign of the future runoff estimate (a 5% uncertainty in ET could yield a positive runoff). Future research needs to focus on reducing this uncertain as noted by Vano et al. (2011).

CO Headwaters Publications MODEL VERIFICATION, SENSITIVITY STUDIES and DOWNSCALING Ikeda, K., R. Rasmussen, C. Liu, D. Gochis, D. Yates, F. Chen, M. Tewari, M. Barlage, J. Dudhia, W. Yu, K. Miller, K. Arsenault, V. Grubišić, G. Thompson, E. Gutmann, 2010: Simulation of seasonal snowfall over Colorado. Atmos. Res. 97, 462-477. Barlage, M., F. Chen, M. Tewari, K. Ikeda, D. Gochis, J. Dudhia, R. Rasmussen, B. Livneh, M. Ek, and K. Mitchell 2010: Noah land surface model modifications to improve snowpack prediction in the Colorado Rocky Mountains. J. Geophys. Res., 115, D22101, doi:10.1029/2009jd013470. Liu, C., K. Ikeda, G. Thompson, R. Rasmussen, and J. Dudhia, 2011: High-resolution simulations of wintertime precipitation in the Colorado Headwaters region: sensitivity to physics parameterizations. Mon. Wea. Rev., doi:10.1175/mwr-d-11-00009.1 Gutmann, E.D., R. M. Rasmussen, C.Liu, K. Ikeda, D. J. Gochis, M.P. Clark, J. Dudhia, G. Thompson, 2011: A comparison of statistical and dynamical downscaling of winter precipitation over complex terrain, In Press. J. Climate. FUTURE CLIMATE COLD SEASON PRECIPITATION STUDY Rasmussen, R., K. Ikeda, C. Liu, D. Gochis, D. Yates, F. Chen, M. Tewari, M. Barlage, J. Dudhia, W. Yu, K. Miller, K. Arsenault, V. Grubišić, G. Thompson, E. Gutmann, 2011: High-Resolution Coupled Climate Runoff Simulations of Seasonal Snowfall over Colorado: A Process Study of Current and Warmer Climate. J. Climate, 24, 3015-3048. FUTURE CLIMATE RUNOFF STUDY Gochis, D., R., K. Ikeda, C. Liu, R. M. Rasmussen, D. Yates, F. Chen, M. Tewari, M. Barlage, J. Dudhia, V. Grubišić, G. Thompson, E. Gutmann (2011): Simulation of seasonal runoff over the Colorado Headwaters regions using high-resolution coupled climate simulations (in preparation).

Acknowledgements: NCAR Water System Program sponsored program (National Science Foundation funding) NCAR Advanced Scientific Discovery supercomputing allocation (> 500k GAUs) The Colorado Headwaters research team: Roy Rasmussen, David Gochis, Kyoko Ikeda, Changhai Liu, Jimy Dudhia, Fei Chen, Mukul Tewari, Mike Barlage, Greg Thompson, Ethan Gutmann, Kristi Arsenault, Vanda Grubisic, David Yates, Kathy Miller UCAR Confidential and Proprietary. 2008, University Corporation for Atmospheric Research. All rights reserved.

Contacts Roy Rasmussen: (rasmus@ucar.edu) DATA and ANALYSIS QUESTIONS Kyoko Ikeda (kyoko@ucar.edu), CC:Roy Rasmussen (rasmus@ucar.edu) WRF MODEL Changhai Liu (liu@ucar.edu) WEAP TOOL and SOCIAL IMPACTS David Yates (yates@ucar.edu) HYDROMETEORLOGY Dave Gochis (gochis@ucar.edu)

Potential New Project with Thai Scientists 1. Conduct PGW type simulations at 4 km grid spacing in order to investigate the impacts of climate change and urbanization on: a. Changes in precipitation, evaporation and runoff b. Floods and drought frequency c. Agricultural production (food security) Note: 4 km resolution allows for the explicit simulation of atmospheric convection versus the typically biased approach using convective parameterizations UCAR Confidential and Proprietary. 2008, University Corporation for Atmospheric Research. All rights reserved.

Second Potential New Project with Thai Scientists 2. Develop seamless minutes to seasons streamflow prediction system a. Combine short term radar nowcasting with high resolution WRF forecasts out to 2 weeks with statistical seasonal forecasts of the meteorology and hydrology. b. Short term forecast is deterministic while long term seasonal prediction is statistical (small increases in skill can have a positive impact on risk management). c. Impacts food security and flood preparations. UCAR Confidential and Proprietary. 2008, University Corporation for Atmospheric Research. All rights reserved.

Common Theme in Projects: WRF-Hydro - My group is integrating hydrological models into WRF to enable full simulation of the hydrological cycle including streamflow - Training and tutorials on WRF-Hydro will be held annually in the summer - Useful for both weather and climate time scales UCAR Confidential and Proprietary. 2008, University Corporation for Atmospheric Research. All rights reserved.