Eric Salathé Climate Impacts Group (JISAO/SMA) University of Washington Constructing regional climate change scenarios.

Eric SalathéClimate Impacts Group

(JISAO/SMA)University of Washington

Constructing regional climate change scenarios

What do Global Models do?

Simulate the global circulation patterns

Can be derived from observations for verification

Precipitation and Temperature is what would occur with global circulation and smooth Earth Does not correspond to observations

Local Scaling

Effect of Wind PatternsWind from South Wind from Southwest

WeakRainshadow

StrongRainshadow

Monthly Flow in Yakima River

Downscaled ECHAM4 IPCC A2-121.7 47.5 (Snoqualmie Pass)

€

PΔ (t) = Phistoric(t)׈ P CC

ˆ P CTR

=σ historic(t)׈ P historic

ˆ P CC

ˆ P CTR

€

PDS ( t) = Pcc(t)׈ P historic

ˆ P CTR

= σ CC (t)׈ P historic

ˆ P CC

ˆ P CTR

“Delta” method -- scale the historic data by the climate-change signal:

Statistical downscaling,-- remove bias in climate-changePrecipitation using the historic data:

The choice is where the time-series characteristicsare taken from:History or Model?

Percent of years with precipitation In bottom 20% of Historic Range

What is the Climate Signal?

Regional Climate Models

Probably not best primary tool for impact studies

Several research issues1. “Second Order” effects

Cold air outbreaks marine layer moves inland.

2. Orographic precipitation response to warming

3. ENSO response

Comparison of RCM and Statistical DownscalingOf Reanalysis GCM

Arbitrary point onCascade Crest

Regional Climate Models

Probably not best primary tool for impact studies

Several research issues1. “Second Order” effects

Cold air outbreaks marine layer moves inland.

2. Orographic precipitation response to warming

3. ENSO response

Summary

1/8-degree P, Tmin, Tmax for IPCC scenarios

What information should we take from models?

What information should we take from historic

record?

What can we (should we) learn from regional

climate models?