Diagnosing and quantifying uncertainties of the reanalyzed clouds, precipitation and radiation budgets over the Arctic and SGP using combined surface-satellite observations

This proposal is in response to the FY13 call of NOAA-OAR-CPO-2013-2003445, MAPP-Research to Advance Climate Reanalysis, with particular focus on addressing some outstanding issues in atmospheric reanalysis. We will use an innovative diagnostic method to quantify the uncertainties of reanalyzed cloud-precipitation-radiation over the Arctic and US SGP regions. The five reanalyses being evaluated in this study are (i) CFSR, (ii) 20CR, (iii) MERRA, (iv) ERA-I, and (V) JRA-25. We will compare the reanalyzed results with observations, find their similarities and differences, and finally investigate how these differences relate to large-scale dynamic patterns and variables using the Self Organizing Maps (SOM) method. The goal of this study is to guide improvement for the cloud/radiation/precipitation parameterizations in these five reanalyses, modify current ones, and/or develop new ones. Therefore, we propose the following two objectives.

Objective 1: Quantifying the uncertainties of reanalyzed Arctic cloud-radiation properties
Reanalyzed cloud properties, such as cloud macrophysical (total/low/middle/high cloud fractions and vertical distribution) and microphysical properties (particle size, LWP/IWP, optical depth) will be evaluated with NASA CERES Ed4 and CloudSat/CALIPSO results over entire Arctic region and locally at the ARM NSA ground-based observations. The reanalyzed surface and TOA radiation fluxes will be evaluated with NASA CERES EBAF results and ARM/BSRN observations. The reanalyzed cloud properties will be compared with CERES-MODIS retrievals for the period 2000-2011. Then we will identify key discrepancies in the comparison for different regions and seasons. To discriminate between these potential problems, Self Organizing Maps (SOM) will be used to classify the atmospheric state from the reanalyses. Intercomparisons of classified atmospheric states between different reanalyses will allow us to determine how the reanalyzed cloud-radiation biases vary with model dynamics.

Objective 2: Investigating the reanalyzed hydrological cycle at SGP
There are two sub-objectives in Objective 2. In the first sub-objective, we will compare the monthly mean cloud fractions and accumulated precipitations from five selected reanalyses with different observational platforms, such as ARM, Oklahoma mesonet system, GPCP, NEXRAD Q2, TRMM, and UND hybrid radar/GOES product over the SGP region during the period 1997-2011 (as we did in Figs 3 and 4). Through this comparison, we will statistically evaluate the strength and weakness in cloud fractions, precipitation strength, frequency occurrence, and areal coverage for each reanalysis. In the second sub-objective, we will perform the case studies during the MC3E over the ARM SGP site during April-June 2011. The aircraft in-situ measured cloud properties, various radar observations and retrievals, and UND hybrid dataset during the MC3E field experiment provide a “cloud/precipitation-truth” dataset and are an invaluable data source for us to evaluate the reanalyzed clouds and precipitation over the SGP region. It is a great addition to the statistical comparison in the first sub-objective.

The two proposed objectives build on our experience in evaluating both GCM/SCM simulations and reanalyses using both surface-satellite observations, and are a natural extension of our current research. They are strongly relevant to one of the NOAA NGSP goals: Assessments of current and future states of the climate system that identify potential impacts and inform science, service, and stewardship decisions. In particular, the two proposed objectives fit in the following MAPP targeted areas: 1) the hydrological cycle, 2) the quality and uncertainties of reanalyses over the Arctic regions, 3) Representation of surface fluxes.