jgrd51861-sup-0001-Readme

advertisement
Supplementary Information for
SPEEDY-IER: A Fast Atmospheric GCM with Water Isotope Physics
Sylvia Dee,1 David Noone,2 Nikolaus Buenning,1
Julien Emile-Geay,1 and Yuxin Zhou1
(1. Department of Earth Sciences, University of Southern California, Los Angeles,
California, USA).
(2. Department of Atmospheric and Oceanic Sciences and Cooperative Institute
for Research in Environmental Sciences, University of Colorado, Boulder, CO,
USA).
[Journal of Geophysical Research: Atmospheres]
Introduction:
This Supplementary Material is designed to provide a complete description of
SPEEDY-IER, including the original model physics, a description of the added
model routines, and relevant equations used in the formulation of added water
isotope physics. Validation metrics standard for isotope-enabled models, but
which are not included in the main text, are reported here, along with additional
sensitivity testing results and model climatology analysis. Figures showing
additional validation and model framework are also provided, as described
below.
1. text01.pdf (Auxiliary Information Text): The Supplementary Material text
describes the original model (SPEEDY) and its physics, summarizes the
changes made to SPEEDY to build SPEEDY-IER, provides an investigation
of how the newly added physics altered the original model’s climatology, and
gives additional model validation metrics and sensitivity tests.
2 List of Auxiliary Figures
2.1 Figure SM1 (fs01_schematic). Schematic of Isotopic Functions and
Subroutines added to SPEEDY. Original structure of SPEEDY; locations that
have been altered are shown in grey, and additions are shown in blue. See text
for explanation of each addition by number.
2.2 Figure SM 2 (fs02_landmodel). Schematic of the Land Model Employed in
SPEEDY-IER Soil water isotope ratios are stored in both the upper surface layer
(RS) and the lower bulkwater layer (RB). Exchange between soil water and the
atmosphere through transpiration is assumed to be non-fractionating. A standard
Craig-Gordon model for evapotranspiration (RET) is assumed during evaporation
from soils. The model assumes there is no fractionation during sublimation.
2.3 Figure SM 3 (fs03_T_bias_evolution). Physics Changes to SPEEDY-IER:
Stepwise Temperature (C). Tempera- ture Biases for 1980-1990 as compared to
NCEP Reanalysis. Original model. (top right): Rain Model added. (bottom left):
Coupled Land Model added. (bottom right): Full Model (land, rain models added).
2.4 Figure SM 4 (fs04_T_bias_difference). Physics Changes to SPEEDY-IER:
Stepwise Temperature Biases (C). Temperature Biases for 1980-1990 as
compared to NCEP Reanalysis. (left): Rain model - Orig- inal model. (center):
Land model-Rain model. (right): Full Model (land, rain models added)- Original.
2.5 Figure SM 5 (fs05_P_bias_evolution). Physics Changes to SPEEDY-IER:
Changes in Precipitation Biases (mm/day). Precipitation Biases. (top left):
Original model. (top right): Rain Model added. (bottom left): Coupled Land Model
added. (bottom right): Full Model (land, rain models added).
2.6 Figure SM 6 (fs06_P_bias_difference). Physics Changes to SPEEDY-IER:
Stepwise Difference in Precipitation Biases (mm/day). Precipitation Biases. (left):
Rain model - Original model. (center): Land model-Rain model. (right): Full Model
(land, rain models added)-Original.
2.7 Figure SM 7 (fs075_colocated_d18o). SPEEDY-IER isotopes in precipitation
compared to GNIP data. Annually averaged δ18OP for model vs. obs, co-located
model points.
2.8 Figure SM 8 (fs07_DJF-JJA_new). SPEEDY-IER simulated seasonal cycle
for isotopes in precipitation. DJF-JJA δ18OP
2.9 Figure SM 9 (fs08_N15_GNIP_dxs). Deuterium excess in precipitation:
annual averages of deuterium ex- cess simulated in a 10-year run of SPEEDYIER compared to the GNIP database. Contours: SPEEDY-IER. GNIP: filled dots.
2.10 Figure SM 10 (fs09_d18OvT_N15). SPEEDY-IER simulated vs. GNIP
δ18OP dependence on Temperature. (a). Scatter plot of δ18OP -temperature
dependence for SPEEDY-IER corresponding to GNIP station (red) and GNIP
observations (green), as well as all SPEEDY-IER simulated data (blue) for 19632000, annually averaged. Reduced dependence on temperature at upwards of
20◦ (due to a shift towards dependence on precipitation amount) is correctly
modeled by SPEEDY-IER. However, across the midlatitudes and poles,
SPEEDY-IER values tend towards more enriched values of δ18OP compared to
the GNIP observations. Best fit lines are calculated for temperatures below 10◦C.
(b). Scatter plot of δ18OP -temperature dependence for SPEEDY-IER
corresponding to GNIP station points (blue) and GNIP station data (green),
annual averages for latitudes poleward of 30◦.
2.11 Figure SM 11 (fs10_N15_GMWL). SPEEDY-IER and the Global Meteoric
Water Line (GMWL) Comparison of SPEEDY-IER δ18O vs δD to the expected
Global Meteoric Water Line (GMWL). SPEEDY- IER comes close to the
expected slope and intercept (8:8.7, 10: 12).
2.12 Figure SM12 (fs11_sciamachy). Deuterium Isotopes in Vapor: SPEEDY-IER
- SCIAMACHY [Model- Observations]. Annual mean difference of δD values in
total column water vapor. The difference map reveals that SPEEDY-IER is in
general agreement with SCIAMACHY values across the tropics, but diverges
towards more enriched values of isotopic vapor moving poleward.
2.13 Figure SM 13 (fs12_SWING_vapor_n15). Evaluation of Performance:
Comparison to other isotope enabled GCMs from the SWING2 runs, Annually
Averaged δ18O� vapor in the surface (925 hPa) layer. (a) SPEEDY-IER δ18O�
vapor (925 hPa), annual average 1970-1989, (b) CAM, (c) LMDZ (free-running),
(d) ECHAM, (e) GISS (free-running), (f) MIROC (free-running)
2.14 Figure SM 14 (fs13_NINO). SPEEDY-IER Correlation with DJF Nino3.4
Surface Temperatures. Global correlation map for SPEEDY-IER simulated
surface air temperatures in the Nino3.4 region and δ18OP , (2-7 year bandpass
filtered monthly data for 1870 to 2000). (a). Covariance of NINO3.4 temperatures
and δ18OP . (b). Correlation of NINO3.4 temperatures and δ18OP . (c). R2
values (% of Described Variance) for δ18OP vs NINO3.4 Temperature anomalies.
2.15 Figure SM 15 (fs14_KE_RMS) SPEEDY-IER RMS errors for KE. Plotted
RMSE for land and ocean points, δ18OP, dxs, GNIP vs. SPEEDY-IER for 19651990
2.16 Figure SM 16 (fs15_KE_RMS_RAIN). SPEEDY-IER RMS errors for KE.
Plotted RMSE for land and ocean points, GPCP Precipitation Rate (mm/day) vs.
SPEEDY-IER for 1965-1990
2.17 Figure SM 17 (fs16_PHI_RMS). SPEEDY-IER RMS errors for φ. Plotted
RMSE for land and ocean points, δ18OP , dxs, GNIP vs. SPEEDY-IER for 19651990
2.18 Figure SM 18 (fs17_fT_RMSE). SPEEDY-IER RMS errors for fT. Plotted
RMSE for land points, δ18OP , dxs, GNIP vs. SPEEDY-IER for 1965-1990
2.19 Figure SM 19 (fs18_KE_PRECIP). The Effects of KE on SPEEDY-IER
Climatology: Experiments-Control. (left) 0x KE precipitation (mm/day), (center)
1xKE precipitation (mm/day), (right) 2x KE precipitation (mm/day)
2.20 Figure SM 20 (fs19_bucket). Isotope ratios in soil water and precipitation:
changes induced by surface bucket depth (bs) variation. Top row (a): δ18O� in
soil water, difference from control. Left: bs=2cm. Center: bs=5cm. Right:
bs=20cm. Bottom row (b): δ18O in precipitation, difference from control. Left:
bs=2cm. Center: bs=5cm. Right: bs=20cm.
Download