Storyboard: Benefits and added value of convection-permitting climate modeling over Fenno-Scandinavia

Benefits and added value of convection-permitting climate modeling over Fenno-Scandinavia

Snow, ice, the long coastline, and the Scandes mountains, they all make the Fenno-Scandinavia a complex region. The box plots show observed distribution of daily mean 2 meter temperature (T2m, top panel) and precipitation (Pr, bottom) for winter (DJF, left half) and summer (JJA, right half) seasons.

A mini-ensemble

A mini-ensemble is used to explore the climate in this region. The modelling region is illustrated in the left figure and the configuration is listed below:

• HCLIM-ALADIN @ 12km resolution, HCLIM-AROME @ 3km
• ERA-I driven runs; 1998–2018
• GCM driven runs; 1985–2005 / 2040–2060 / 2080–2100
• 2 forcing GCMs: EC-Earth & GFDL-CM3 (CMIP5)
• 2 emission scenarios: RCP8.5 and RCP4.5

Why the use of CPM model?

• For precipitation, there are obvious arguments in favor of using CPM models over Fenno-Scandinavia (and elsewhere).
• Left panel: Probability density function (PDF) of summer (JJA) hourly precipitation over Sweden comparing HCLIM3 CPM and HCLIM12 regional climate model (RCM) with radar observations.
• Right panel: Maps of 99.9th percentile of summer (JJA) hourly precipitation over Sweden in HCLIM12, HCLIM3 and radar observations.

More precipitation extremes in scenarios

• End-of-C (2080-2100), RCP8.5 changes in precipitation distributions;
• Map plots: CPM change signal in average (left) and heavy (right) summer (JJA) precipitation,
• Line plots: ASoP analysis for present climate (top) and changes (scn - ctrl) (bottom) of summer (JJA) hourly precipitation (Pr).
• Distinct shifts in precipitation distributions - reductions in low intensities, increases in higher intensities.
• More emphasized change in CPM compared to RCM.

More precipitation extremes in scenarios

• Changes (in % per degree warming) of high precipitation (Pr) percentiles (hourly data) in winter and summer over Fenno-Scandinavia.
• Stronger GCM influence in winter when RCM/CPM combinations are relatively close.
• Differences in RCM and CPM responses are much more pronounced in summer when forcing from boundaries is not as constraining (compared to DJF) and local convection becomes more prominent.

Running to the hills...

• Ambition to set focus on the Scandes mountains; if and how CPM may provide added value.
• These figures show changes in snow climate variables (x axes) as a function of altitude (y axes), by end-of-C in RCP4.5 and 8.5.
• Solid Precipitation (top left), solid Precipitation/total Precipitation ratio (top right), snow cover (bottom left), # days with large snow depth (bottom right).

Orographic precipitation in winter

• Separating windward/leeward model responses in the Scandes indicate systematic differences between CPM and RCM.
• If a consistent response it will be an important aspect of expected regional climate change in Scandes.
• Further investigation needed.

Not just precipitation?

• Added value from CPM is evident for precipitation - what about other variables?
• We could expect improved representations of winds (e.g. convective wind gusts, orographic down-slope wind storms, thermal circulations).

Figures show ongoing work:

• Top panels: PDFs of winter near surface wind speed, comparing CPM (red) and RCM (blue) with station values in Sweden (black). Left panel is for low-lying stations and right panel for stations in mountains.

In mountains there is clear indication of added value in CPM model, especially in the tail distribution.

• Bottom figure shows the wind speed as a function of altitude in present climate in RCM (blue/green) and CPM (red/purple).

How do we benefit from using CPM models in Fenno-Scandinavia?

• More realistic representation of precipitation, especially in the summer season (improvements in frequency, duration and intensity)
• Potential added value through systematically different climate change response in heavy summer precipitation.
• Some suggestions of different climate change responses between CPM and RCM in solid versus liquid precipitation in the Scandes mountains.
• Near-surface wind speeds are also better represented in the Scandes. Ongoing work to investigate potential added value in climate scenarios as well as more process based perspectives (e.g. orographic down-slope wind-storms).