K. R. Sperber, H. Annamalai, I.-S. Kang,
A. Kitoh, A. Moise, A. Turner, B. Wang, and T. Zhou
Author Version of the Published Journal Article
Climate Dynamics Version of the Published Journal Article
Sperber, K. R.; Annamalai, H.; Kang, I.-S.; Kitoh, A.; Moise, A.; Turner, A.; Wang, B.; Zhou, T., 2012, "The Asian summer monsoon: an intercomparison of CMIP5 vs. CMIP3 simulations of the late 20th century", Climate Dynamics, DOI: 10.1007/s00382-012-1607-6
Click on a diagnostic heading (2nd row) to get a .pdf file of the full suite of observed and model figures and/or skill score plot for that diagnostic.
Table 2: Skill scores for the June-September climatology and the climatological annual cycle. The results are given for observations, the MMM’s, and for the CMIP5 and CMIP3 models. The observed skill for precipitation is between GPCP and CMAP, and the skill for the 850hPa wind (850hPa) is between ERA40 and JRA25. The model pattern correlations for the precipitation climatology (Pr) are calculated with respect to GPCP precipitation. For the 850hPa wind climatology (850hPa), the model pattern correlations are calculated with respect to ERA40 850hPa wind. For the climatologies the skill is calculated over the region 40oE-160oE, 20oS-50oN. For the time-latitude (T-Lat) climatological annual cycle of monthly rainfall averaged between 70oE-90oE, the model pattern correlations are calculated with respect to GPCP precipitation over the region 10oS-30oN, for May-October (see Section 4.1). For the climatological annual cycle of pentad rainfall, the model pattern correlations are calculated with respect to GPCP precipitation for the pentads of onset, peak, withdrawal, and duration of the monsoon over the region 50oE-180oE, 0o-50oN (see Section 4.2). The categorical skill scores, hit rate and threat score, indicate how well a model represents the spatial domain of the monsoon, where a value = 1 indicates perfect agreement between model and observations. Missing table entries occur for models that did not have available data for analysis. The top five models with the largest skill scores for each diagnostic are highlighted
Table 2
Model |
Climatology |
Climatological Annual Cycle Rainfall |
|||||||
|---|---|---|---|---|---|---|---|---|---|
Observations |
0.927 |
0.986 |
0.887 |
0.748 |
0.834 |
0.830 |
0.671 |
0.893 |
0.744 |
CMIP5 MMM |
0.898 |
0.976 |
0.674 |
0.664 |
0.786 |
0.792 |
0.605 |
0.844 |
0.625 |
CMIP3 MMM |
0.865 |
0.967 |
0.657 |
0.510 |
0.733 |
0.712 |
0.380 |
0.821 |
0.573 |
BCC-CSM-1 |
0.808 |
0.928 |
0.338 |
||||||
bccr-bcm2.0 |
0.733 |
0.933 |
0.639 |
||||||
CanESM2 |
0.815 |
0.951 |
0.552 |
0.298 |
0.451 |
0.543 |
0.164 |
0.782 |
0.517 |
cgcm3.1 (t47) |
0.782 |
0.935 |
0.465 |
0.063 |
0.476 |
0.454 |
0.109 |
0.766 |
0.522 |
cgcm3.1 (t63) |
0.796 |
0.944 |
0.461 |
0.155 |
0.432 |
0.384 |
0.154 |
0.758 |
0.508 |
CCSM4 |
0.849 |
0.952 |
0.678 |
0.581 |
0.717 |
0.798 |
0.570 |
0.836 |
0.619 |
ccsm3 |
0.748 |
0.913 |
0.390 |
0.394 |
0.481 |
0.459 |
0.346 |
0.757 |
0.487 |
pcm1 |
0.634 |
0.793 |
0.364 |
||||||
CNRM-CM5 |
0.852 |
0.974 |
0.567 |
0.674 |
0.638 |
0.750 |
0.656 |
0.796 |
0.513 |
cnrm-cm3 |
0.717 |
0.908 |
0.763 |
0.489 |
0.596 |
0.633 |
0.329 |
0.749 |
0.437 |
CSIRO-Mk3.6.0 |
0.713 |
0.896 |
0.232 |
0.006 |
0.451 |
0.729 |
0.331 |
0.762 |
0.497 |
csiro-mk3.0 |
0.803 |
0.889 |
0.385 |
0.196 |
0.461 |
0.601 |
0.147 |
0.790 |
0.495 |
csiro-mk3.5 |
0.796 |
0.923 |
0.171 |
0.287 |
0.474 |
0.665 |
0.350 |
0.788 |
0.540 |
FGOALS-g2 |
0.766 |
0.923 |
0.455 |
||||||
FGOALS-s2 |
0.807 |
0.916 |
0.613 |
0.601 |
0.596 |
0.649 |
0.531 |
0.812 |
0.537 |
fgoals-g1.0 |
0.690 |
0.803 |
0.587 |
-0.050 |
0.672 |
0.785 |
0.097 |
0.770 |
0.460 |
GFDL-CM3 |
0.844 |
0.941 |
0.742 |
0.458 |
0.407 |
0.546 |
0.406 |
0.796 |
0.532 |
GFDL-ESM2G |
0.821 |
0.955 |
0.727 |
0.370 |
0.560 |
0.660 |
0.328 |
0.841 |
0.615 |
GFDL-ESM2M |
0.828 |
0.958 |
0.676 |
0.490 |
0.714 |
0.730 |
0.383 |
0.824 |
0.586 |
gfdl_cm2_0 |
0.826 |
0.954 |
0.673 |
0.715 |
0.540 |
0.624 |
0.495 |
0.812 |
0.559 |
gfdl_cm2_1 |
0.843 |
0.957 |
0.681 |
0.453 |
0.662 |
0.731 |
0.485 |
0.825 |
0.587 |
GISS-E2-H |
0.631 |
0.902 |
0.318 |
||||||
GISS-E2-R |
0.730 |
0.912 |
0.235 |
||||||
giss_aom |
0.780 |
0.894 |
0.282 |
0.359 |
0.614 |
0.540 |
0.203 |
0.774 |
0.457 |
HadCM3 |
0.773 |
0.931 |
0.550 |
0.555 |
0.447 |
0.519 |
0.452 |
0.873 |
0.675 |
HadGEM2-CC |
0.795 |
0.927 |
0.376 |
0.526 |
0.659 |
0.634 |
0.317 |
0.777 |
0.543 |
HadGEM2-ES |
0.800 |
0.933 |
0.356 |
0.562 |
0.620 |
0.648 |
0.367 |
0.769 |
0.538 |
ukmo_hadcm3 |
0.778 |
0.932 |
0.529 |
||||||
ukmo_hadgem1 |
0.798 |
0.938 |
0.386 |
||||||
ingv-sxg |
0.814 |
0.950 |
0.629 |
0.277 |
0.575 |
0.724 |
0.417 |
0.797 |
0.516 |
INMCM4 |
0.742 |
0.864 |
0.561 |
0.153 |
0.616 |
0.649 |
0.224 |
0.810 |
0.560 |
inmcm3.0 |
0.619 |
0.837 |
0.497 |
-0.125 |
0.331 |
0.592 |
-0.064 |
0.795 |
0.517 |
IPSL-CM5A-LR |
0.797 |
0.926 |
0.442 |
0.399 |
0.540 |
0.712 |
0.482 |
0.798 |
0.515 |
IPSL-CM5A-MR |
0.809 |
0.935 |
0.501 |
0.421 |
0.575 |
0.769 |
0.591 |
0.787 |
0.501 |
ipsl-cm4 |
0.743 |
0.907 |
0.214 |
0.215 |
0.495 |
0.634 |
0.254 |
0.786 |
0.468 |
MIROC-ESM |
0.617 |
0.824 |
0.518 |
0.391 |
0.610 |
0.666 |
0.394 |
0.756 |
0.434 |
MIROC-ESM-CHEM |
0.642 |
0.831 |
0.538 |
0.518 |
0.669 |
0.653 |
0.423 |
0.752 |
0.433 |
MIROC4h |
0.802 |
0.940 |
0.573 |
0.674 |
0.626 |
0.766 |
0.620 |
0.843 |
0.611 |
MIROC5 |
0.842 |
0.940 |
0.778 |
0.362 |
0.778 |
0.851 |
0.652 |
0.808 |
0.531 |
miroc3.2(hires) |
0.761 |
0.914 |
0.523 |
0.483 |
0.383 |
0.709 |
0.568 |
0.792 |
0.486 |
miroc3.2(medres) |
0.765 |
0.919 |
0.513 |
0.633 |
0.402 |
0.571 |
0.503 |
0.744 |
0.384 |
MPI-ESM-LR |
0.792 |
0.949 |
0.664 |
0.316 |
0.579 |
0.652 |
0.472 |
0.781 |
0.535 |
echam5/mpi-om |
0.800 |
0.942 |
0.664 |
0.265 |
0.412 |
0.537 |
0.337 |
0.800 |
0.547 |
miub_echo_g |
0.803 |
0.911 |
0.522 |
0.008 |
0.041 |
0.368 |
0.189 |
0.787 |
0.507 |
MRI-CGCM3 |
0.752 |
0.886 |
0.195 |
0.024 |
0.619 |
0.535 |
-0.014 |
0.751 |
0.465 |
mri-cgcm2.3.2 |
0.726 |
0.885 |
0.538 |
0.471 |
0.345 |
0.550 |
0.346 |
0.746 |
0.473 |
NorESM1-M |
0.848 |
0.913 |
0.634 |
0.558 |
0.723 |
0.791 |
0.565 |
0.838 |
0.624 |
Table 3: Skill scores for the Indian Monsoon and East Asian Monsoon interannual variability and the boreal summer intraseasonal variability (BSISV). The results are given for observations, the MMM’s, and for the CMIP5 and CMIP3 models. The interannual variations of the ENSO-Monsoon relationship are characterized by (1) the lag 0 correlation between JJAS anomalies of all-India rainfall and NINO3.4 SST (AIR/N3.4). The AIR is for land-only gridpoints over the region 65oE-95oE, 7oN-30oN. The observations are for the anomalies of Rajeevan rainfall vs. HadISST SST for 1961-1999, and (2) the pattern correlations of JJAS precipitation anomalies (Pr) obtained from regression with JJAS anomalies of NINO3.4 SST. The model pattern correlations are calculated with respect to GPCP anomalies that were obtained by regression with the NINO3.4 SST anomalies from the NCEP/NCAR reanalysis (1979-2007). The pattern correlations are calculated over the region 60oE-100oE, 0o-30oN. For observations the skill is between GPCP and CMAP. For the East Asian Monsoon, the negative of the June-August Wang and Fan (1999) zonal wind shear index (WFN, see Section 5.2) is regressed against June-August anomalies of precipitation and 850hPa wind. The model pattern correlations are calculated with respect to GPCP rainfall anomalies and JRA 850hPa wind anomalies that were obtained by regression with the JRA25 WFN. The pattern correlations are calculated over the region 100oE-140oW, 0o-50oN. For observations the skill is between GPCP/JRA25 and CMAP/NCEP-NCAR Reanalysis. For BSISV, the skill is (1) the pattern correlation of June-September 20-100 day bandpass filtered OLR variance between the model (1961-1999) and AVHRR OLR (1979-2006). For observations the skill is for AVHRR OLR for 1979-2006 vs. AVHRR OLR for 1979-1995, and (2) the spatio-temporal correlation of the model BSISV life cycle vs. that from the observed cyclostationary EOF (CsEOF) analysis of Annamalai and Sperber (2005). The life cycle of the BSISV is obtained by first projecting 20-100 day filtered OLR from observations (1979-2006) and the models (1961-1999) on to the Day 0 pattern of the observed CsEOF. The resulting PC is used for lag regression against the 20-100 day filtered OLR with the spatio-temporal correlation between model and observation being calculated for Day -15, Day -10, Day -5, Day 0, Day 5, Day 10, Day 15, and Day 20. The skill scores for the intraseasonal variability are calculated over the region 40oE-180oE, 30oS-30oN. Missing table entries occur for models that did not have available data for analysis. The top five models with the largest skill scores for each diagnostic are highlighted
Table 3
Model |
Indian Monsoon |
East Asian Monsoon |
BSISV |
|||
|---|---|---|---|---|---|---|
Observations |
-0.533 |
0.798 |
0.959 |
0.989 |
0.995 |
0.893 |
CMIP5 MMM |
0.616 |
0.888 |
0.972 |
0.903 |
0.766 |
|
CMIP3 MMM |
0.600 |
0.799 |
0.969 |
0.895 |
0.754 |
|
BCC-CSM-1 |
-0.250 |
-0.140 |
0.695 |
0.930 |
||
bccr-bcm2.0 |
-0.430 |
0.249 |
0.670 |
0.951 |
||
CanESM2 |
-0.273 |
0.014 |
0.672 |
0.861 |
0.846 |
0.651 |
cgcm3.1 (t47) |
-0.335 |
0.404 |
0.625 |
0.899 |
0.727 |
0.605 |
cgcm3.1 (t63) |
-0.182 |
0.173 |
0.703 |
0.938 |
0.717 |
0.604 |
CCSM4 |
-0.556 |
0.337 |
0.789 |
0.947 |
||
ccsm3 |
-0.561 |
0.264 |
0.722 |
0.800 |
0.695 |
0.588 |
pcm1 |
-0.356 |
0.293 |
0.232 |
0.870 |
||
CNRM-CM5 |
-0.307 |
0.245 |
0.642 |
0.894 |
||
cnrm-cm3 |
-0.484 |
0.419 |
0.313 |
0.727 |
0.570 |
0.600 |
CSIRO-Mk3.6.0 |
-0.487 |
0.162 |
0.346 |
0.858 |
0.809 |
0.645 |
csiro-mk3.0 |
-0.403 |
-0.112 |
0.629 |
0.939 |
0.830 |
0.581 |
csiro-mk3.5 |
-0.719 |
0.137 |
0.569 |
0.924 |
||
FGOALS-g2 |
-0.052 |
0.238 |
0.739 |
0.936 |
||
FGOALS-s2 |
0.114 |
0.096 |
0.787 |
0.921 |
0.734 |
0.608 |
fgoals-g1.0 |
-0.747 |
0.276 |
0.415 |
0.426 |
0.271 |
0.438 |
GFDL-CM3 |
-0.442 |
0.192 |
0.315 |
0.867 |
||
GFDL-ESM2G |
-0.289 |
0.251 |
0.458 |
0.972 |
0.753 |
0.643 |
GFDL-ESM2M |
-0.187 |
0.251 |
0.606 |
0.955 |
||
gfdl-cm2.0 |
-0.667 |
0.336 |
0.668 |
0.976 |
0.818 |
0.677 |
gfdl-cm2.1 |
-0.494 |
0.412 |
0.390 |
0.919 |
0.850 |
0.712 |
GISS-E2-H |
-0.094 |
0.254 |
0.586 |
0.918 |
||
GISS-E2-R |
-0.366 |
0.379 |
0.656 |
0.906 |
||
giss-aom |
0.094 |
0.189 |
0.117 |
0.754 |
-0.070 |
0.395 |
HadCM3 |
-0.299 |
0.180 |
0.773 |
0.897 |
||
HadGEM2-CC |
-0.335 |
-0.068 |
0.787 |
0.935 |
0.857 |
0.641 |
HadGEM2-ES |
-0.344 |
0.216 |
0.839 |
0.949 |
0.862 |
0.651 |
ukmo-hadcm3 |
-0.374 |
0.323 |
0.758 |
0.947 |
||
ukmo-hadgem1 |
-0.446 |
0.154 |
0.744 |
0.912 |
||
ingv-sxg |
-0.455 |
0.313 |
0.513 |
0.925 |
||
INMCM4 |
-0.033 |
0.110 |
-0.047 |
0.816 |
0.639 |
0.562 |
inm-cm3.0 |
-0.258 |
-0.073 |
0.520 |
0.850 |
||
IPSL-CM5A-LR |
-0.700 |
0.611 |
0.450 |
0.708 |
0.791 |
0.654 |
IPSL-CM5A-MR |
-0.763 |
0.636 |
0.532 |
0.749 |
0.827 |
0.635 |
ipsl-cm4 |
-0.554 |
0.347 |
0.675 |
0.787 |
0.785 |
0.648 |
MIROC-ESM |
0.088 |
0.061 |
0.596 |
0.694 |
0.548 |
0.516 |
MIROC-ESM-CHEM |
-0.104 |
0.045 |
0.687 |
0.882 |
0.554 |
0.528 |
MIROC4h |
-0.327 |
0.529 |
0.723 |
0.921 |
0.736 |
0.625 |
MIROC5 |
-0.321 |
0.010 |
0.567 |
0.946 |
0.805 |
0.691 |
miroc3.2(hires) |
0.080 |
-0.009 |
0.643 |
0.915 |
0.666 |
0.543 |
miroc3.2(medres) |
-0.329 |
0.234 |
0.719 |
0.928 |
0.800 |
0.575 |
MPI-ESM-LR |
-0.291 |
0.401 |
0.283 |
0.899 |
0.874 |
0.681 |
echam5/mpi-om |
-0.573 |
0.560 |
0.230 |
0.817 |
0.873 |
0.721 |
miub-echo_g |
-0.554 |
0.113 |
0.664 |
0.914 |
0.810 |
0.702 |
MRI-CGCM3 |
-0.274 |
0.338 |
0.819 |
0.937 |
0.782 |
0.628 |
mri-cgcm2.3.2 |
-0.424 |
0.107 |
0.570 |
0.931 |
0.575 |
0.654 |
NorESM1-M |
-0.690 |
0.522 |
0.811 |
0.959 |
0.833 |
0.627 |
For further information please contact:
Kenneth R. Sperber, Ph.D.
Program for Climate Model Diagnosis and Intercomparison
Lawrence Livermore National Laboratory
P.O. Box 808, L-103
Livermore, CA 94551 USA
Email: sperber1@llnl.gov
Ph: 925-422-7720
Fax: 925-422-7675
Acknowledgements
K. R. Sperber was supported by the Office of Science (BER), U. S. Department of Energy through Lawrence Livermore National Laboratory contract DE-AC52-07NA27344. The authors thank Charles Doutriaux for assistance in making the animations, Renata McCoy for developing the webpage. LLNL-WEB-600275