HiGEM land surface paper

1
HiGEM land surface paper
Douglas Clark
Christopher Taylor
Centre for Ecology and Hydrology, Wallingford
2
Progress
Contribution to HiGEM overview paper
Work in progress Finish in first 2 weeks Feb
Land surface paper
Still at an early stage
3
Outline of land surface paper
Description of climate over land
Look at drivers of the land model precip, T,
radiation.
Snow extent in coupled and uncoupled runs
River flow in coupled and uncoupled runs
Soil moisture stress in coupled and uncoupled runs
Compare with EO product.
Can we relate biases in different fields?
Focus on key regions/catchments
Whats the story? Need to emphasise/develop new
aspects if we want to get published!
4
Precipitation and temperature (on land)
Annual averages
Precipitation (mm d-1)
HiGEM has smaller wet bias all seasons, global
and tropics. gt50N not much difference
HiGEM-HadGEM, wet bias slightly worse in HiGEM.
Temperature (K)
HiGEM has smaller cool bias all seasons, global
and tropics. gt50N smaller cool bias, except JJA
now has larger warm bias.
5
Precipitation (on land)
Wet bias, worse in HiGEM USA, Brazilian
highlands, Uruguay
Wet bias, better in HiGEM Indonesia, Siberia,
southern Africa, central S.America, SW. Europe,
E.China
Dry bias, worse in HiGEM W.Africa
6
Precipitation
50 major catchments with (relatively) long-term
riverflow data.
Exclude those with most obvious flow
regulation/abstraction e.g. Parana.
7
Precipitation
8
GCMs too wet in all of these. Poor offline
suggests model deficiencies.
River flow the four largest catchments
9
River flow
Too much E?
Offline far from perfect!
10
River flow high latitudes
Slightly too little flow, too late. Offline
timing better with old soil.
Issues include Infiltration into partially
frozen soil. Simple runoff model.
11
Moisture stress
Richard Ellis et al.
12
Snow extent
HiGEM
HadGEM
Offline (1deg)
GCMs wet bias (small difficulty with obs) and
cold bias.