Investigating the sea surface temperature representation and its regional climate influence in Southeast Asia
Climate variability over Southeast Asia is affected by variability in sea surfacetemperatures (SST). However, spatial patterns and temporal variability of SST and itsregional influence may not be well represented in climate models, resulting in biases inthe downscaled climate output. Historical clim...
Saved in:
| Main Author: | |
|---|---|
| Format: | text |
| Published: |
Archīum Ateneo
2018
|
| Subjects: | |
| Online Access: | https://archium.ateneo.edu/theses-dissertations/2 http://rizalls.lib.admu.edu.ph/#section=resource&resourceid=1397732173&currentIndex=0&view=fullDetailsDetailsTab |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Institution: | Ateneo De Manila University |
| Summary: | Climate variability over Southeast Asia is affected by variability in sea surfacetemperatures (SST). However, spatial patterns and temporal variability of SST and itsregional influence may not be well represented in climate models, resulting in biases inthe downscaled climate output. Historical climate simulations by Southeast AsiaRegional Climate Downscaling (SEACLID)/Coordinated Regional Climate Downscaling Experiment (CORDEX) Southeast Asia showed biases in temperatureand rainfall, which indicates a need to examine SST representation in climate modelsin order to investigate and address possible associated uncertainties. This study aims todescribe the historical SST over Southeast Asia from observation in terms of spatialpatterns and temporal variability; to analyze the CMIP5 GCM representation of SSTover Southeast Asia and its potential influence on the model climate; and to assess thepossible effect of SST representation in CMIP5 GCMs to the downscaled regionalclimate output.Results show that four GCMs that best represent SSTs also represent climatevariables well (CNRM-CM5, IPSL-CM5A-LR, HadGEM2-AO, and HadGEM2-ES).BCC-CSM1.1 did not simulate temperature, precipitation and wind speed well despiteits well simulated SST. Particular GCMs that do not represent SST well (MRI-CGCM3,MRI-ESM1 and NorESM1-M) also produced climate simulations that are far fromobserved data. We note though that there are GCMs that do not show any relationship between SST representation and the resulting climate simulation. Moreover, in thevSWMA where the SST is simulated well and the seasonal variability is captured w ell,temperature, rainfall and winds are also adequately represented by the models (IPSLCM5A-LR, MPI-ESM-MR, MIROC5). Temperature in particular appears to be thevariable that is closely correlated with SST. Further analysis of climate variables showsthat in both GCMs and downscaled regional models, colder land surface, wetterconditions due to precipitation, and stronger winds, with respect to observation andreanalysis, are associated with colder seas compared to observation.Findings of this research may give a better understanding on how SSTinfluences modeled climatology in Southeast Asia and can help improve regionalclimate model simulations for better future climate projections used for adaptation andimpact studies. |
|---|