Statistical Downscaling of Rainfall for the Hawaiian Islands using CMIP3 Model Scenarios

We applied a statistical downscaling method in order find a connection between the large-scale atmospheric circulation over the Pacific (180-120W 10S-40N) with the rainfall over Hawaii. Using observational records of precipitation at individual rainfall stations across the Hawaiian islands, we first identified the circulation pattern that occurred during months with low and high precipitation at each single station during the last 50 years. In this study we only used the near surface winds. Especially, the strength and direction of the wind in north-south direction (e.g. Trade Winds, Kona Winds) is a useful indicator for low/heavy rainfall months in many parts of the Hawaiian Islands.
We carefully tested how robust the relationships between rainfall and the wind circulation over the Pacific are during the dry summer months (May-October) and the wet winter months (November-April). This important step in the statistical downscaling excluded some stations from our further application (marked with x in the maps). Note that some stations had insufficient observations for a robust statistical evaluation (marked with + in the maps).

Finally, we looked into the future climate change scenarios from the IPCC model simulations. We selected the A1B emission scenario. This scenario is in the middle range of IPCC's projected scenarios of fossil fuel consumption. 6 models were analyzed for their circulation changes over the Pacific by the end of the 21st century (2070-2099). The wind changes were compared with the wind pattern that have been associated with low/high monthly mean rainfall amounts of the past 50 years. We estimated the projected rainfall changes by measuring the similarity among the simulated future wind changes with the past observed wind pattern during the last 50 years.

The maps show the stations and the estimated rainfall changes during the late 21st century in percent of the present average rainfall. We provide the two numbers, one for the dry season (May-October) and one for the wet season (November-April). Further, the rainfall change is expressed in inches per month for each individual station.
We also provide the 'ensemble mean' changes, that is the average projected rainfall changes of all six climate models.


Timm, O. and H. F. Diaz, 2009: Synoptic-Statistical Approach to Regional Downscaling
of IPCC 21st Century Climate Projections: Seasonal Rainfall over the Hawaiian Islands
J. Climate, 22, 4261–4280, doi: 10.1175/2009JCLI2833.1