Introduction of the problem and rationale:
Faced with the prospect of climate change due to global warming, detailed projections of how regional climate may change are needed for policy makers to assess different strategies for adaptation and mitigation. Currently, regional aspects of climate change still faces tremendous uncertainties, both in terms of the ability of global climate models to adequately simulate regional climate, as well as the need of a physically and dynamically consistent long term dataset to assess regional climate variability and to validate climate model simulations. Previous studies, including those by the PI, have shown that previous reanalysis datasets have spurious trends in the statistics of synoptic scale weather systems due to improvements in the observation system over time, and thus are not suitable for the characterization of decadal variability and trends of these systems. The new dataset from the 20th Century Reanalysis Project (20CR) is expected to suffer from less of such time dependent biases because it uses an advanced data assimilation system to assimilate only a single type of observations (surface pressure observations) that had undergone the least changes over time. However, the frequency and quality of observations ingested into 20CR have still undergone significant temporal variations over the 20th Century, thus the quality and consistency of the statistics of synoptic scale weather systems derived from the 20CR still needs to be assessed.
Brief summary of work to be completed:
The statistics of synoptic scale weather systems, in terms of variance/covariance statistics of mean sea level pressure and other quantities such as geopotential height, meridional wind, and temperature, will be derived from 20CR and compared directly to similar statistics derived from surface ship and upper-air radiosonde observations to assess the quality and consistency of such statistics derived from the 20CR. Since variability of synoptic scale weather systems and mean flow variability are closely tied to each other, statistical models will be used to assess whether the relationships between synoptic scale and mean flow variabilities are consistent throughout the 20th Century. Feature tracking statistics (including statistics of surface and upper level cyclones) will be derived from 20CR and compared to those derived from other reanalysis products to assess how changes in the observing system might have affected cyclone statistics. The completion of this project will result in better understanding of past variability of synoptic scale weather systems, and provide climate scientists with an assessment of the period over which synoptic scale variability derived from 20CR is consistent, and climate modelers with a validation of the dataset that they can use to assess climate model simulations, thereby resulting in better understanding of the ability of climate models to simulate and project changes in regional climate, resulting in improved assessment of the projected regional climate change by climate models.