Menu Close


Task 233

Evaluation of the NASA GMAO Modern Era Retrospective-Analysis for Research and Applications (MERRA) in Polar Latitudes

Principal Investigator(s):

R. I. Cullather


M. Bosilovich

Last Updated:

May 14, 2013 16:46:22

Description of Problem

MERRA is a state-of-the-art global numerical reanalyses that has recently been produced by the NASA GMAO, and covers the period from 1979 until the present. Reanalyses are 4-dimensional numerical depictions of the atmospheric state that are retrospectively produced through the assimilation of available observations using a weather prediction model. The objective of MERRA is to provide a climate context for the NASA satellite observing system and to improve the representation of the water cycle in reanalyses. The purpose of this work is to characterize the performance of MERRA over polar regions including the Arctic Ocean and continental ice sheets and, where appropriate, apply MERRA for the purpose of regional climate study. An assessment of MERRA in these regions is useful for providing guidance to other users in locations that are of particular importance in climate research. The characterization of high latitude regional climate variability and its depiction in reanalysis products is also applicable to ongoing model validation efforts at GMAO.

Scientific Objectives and Approach

In the first stage of this work, the focus has been on assessing components of the atmospheric energy and moisture balances with prior studies. To this end a combined approach has been used to compare fixed sub-regions with prior studies, and to make concurrent comparisons with in situ observations at point locations. The second stage is the application of MERRA for regional climate study. The approach used here is to take advantage of lessons learned from the budget assessments in order to understand the strengths and limitations of reanalyses. This work also makes use of other data products including those from the European Centre for Medium-Range Weather Forecasts (ECMWF) and the U.S. National Centers for Environmental Prediction (NCEP) in order to improve confidence in the conditions described by MERRA.


Cullather and Bosilovich (2011) evaluates the polar atmospheric moisture budget in MERRA and has been accepted for publication. A companion paper entitled “The energy budget of the polar atmosphere in MERRA” has been reviewed and is undergoing minor revision. These two papers will appear in special issues of the Journal of Climate on MERRA and on the U.S. CLIVAR/SeaFlux Workshop on Surface Fluxes: Challenges for High Latitudes. Results from the two papers were presented at the High Latitude Surface Fluxes workshop in Boulder, CO, and submission of these manuscripts to the second special issue was invited.

Several applications for regional climate study have been pursued. The first is an analysis of prevalent atmospheric circulation conditions during the recent decrease in perennial sea ice cover in the Arctic. Recent attention has been focused on anomalous conditions during the Arctic summer. As shown in Fig. 1, summertime conditions are historically characterized by low pressure in the central Arctic. This feature arises from baroclinic instability in coastal Eurasia and North America, where relatively warm continents contrast with the cold polar ocean. Low pressure systems form, progress to the north, and undergo cyclolysis in the vicinity of the North Pole. This averaged low pressure feature produced by the storm track is found to have been absent in recent years. Anomalous high pressure reduces the sea ice cover by promoting Ekman convergence, which draws ice out from marginal seas. High pressure also increases sea ice melting with reduced cloudiness and an increased poleward heat transport. By utilizing the hourly, high resolution sea level pressure fields from MERRA, a cyclone tracking algorithm has been developed to understand these circulation changes. Preliminary analysis suggests that summertime baroclinic zones have not changed, but that Eurasian cyclone trajectories have become more zonal in recent years. This could be due to cyclones preferentially passing over newly open water in coastal areas, changes in upstream conditions arising from the loss of sea ice cover near Scandinavia, or for other reasons. Results have been presented at a GMAO internal seminar and at the San Francisco AGU Meeting, and are being assembled for publication. The cyclone tracks developed from this study may be further exploited for investigations of the atmospheric general circulation.

Two other applications take advantage of recent Antarctic field studies. One involves the examination of an Antarctic coastal polynya. A coastal polynya is a region of persistent open water surrounded by pack ice. The polynya in Terra Nova Bay, Antarctica is formed by the confluence of strong katabatic winds which descend off the Antarctic plateau and reach speeds greater than 40 m s−1. This phenomenon modifies local conditions through the exchange of fluxes between the atmosphere and the open water, and influences global oceanic circulation through the production of water masses. In September 2009, colleagues from the University of Colorado at Boulder operated unmanned aircraft over the active polynya. Reanalyses such as MERRA do not capture local features of the confluence zone due to the high spatial resolution required. Nevertheless, MERRA has been employed to document prevailing synoptic conditions that lead to the episodic opening of the polynya. Regression analysis using MERRA and daily AMSR-E sea ice concentration has shown that synoptic high pressure off Adelie Land, East Antarctica is closely related to the polynya opening. This work was presented at a working group meeting in Boulder, CO and is being assembled for publication. Additionally, surface heat fluxes from MERRA are being compared with the observed fluxes from the unmanned aircraft over the polyna. The other project involves the use of a snow particle counter deployed by Columbia University colleagues on research ship cruises in coastal Antarctica over the period 2007-2009. Such measurements are rare and provide a unique opportunity to evaluate MERRA in a data-sparse region. Comparisons indicate that MERRA performs adequately in reproducing the frequency of snow events. These results are also being developed for publication.

Refereed Journal Publications

Cullather, R. I., and M. G. Bosilovich, 2011: The moisture budget of the polar atmosphere in MERRA. J. Climate, in press.

Other Publications and Conferences

Cullather, R. I., 2010: Evaluation of Arctic Energy and Moisture Budgets in the MERRA Reanalysis. US CLIVAR/SeaFlux Workshop on Surface Fluxes: Challenges for High Latitudes, March 17-19, 2010, Boulder, CO.

Cullather, R. I., 2010: Evaluation of MERRA in Polar Regions. GMAO Science Meeting, September 30, 2010, Greenbelt, MD.

Cullather, R. I., and M. G. Bosilovich, 2010: The 21st-Century Arctic Hydrologic Cycle in Reanalyses (poster). U.S. CLIVAR Workshop on the Evaluation of Reanalyses: Developing an Integrated Earth System Analysis (IESA) Capability, November 1-3, 2010, Baltimore, MD.

Cullather, R. I., M. G. Bosilovich, and B. Tremblay, 2010: Atmospheric Circulation in Reanalyses Concurrent With Recent Arctic Sea-Ice Decline. NASA GMAO Internal Seminar Series, November 9, 2010, Greenbelt, MD.

Cullather, R. I., 2010: Terra Nova Bay Polynya in Relation to Atmospheric Forcing, Meeting of the Working Group on Atmosphere-Ocean-Ice Interaction in a Coastal Polynya, November 15, 2010, Boulder, CO.

Cullather, R. I., and M. G. Bosilovich, 2010: MERRA Arctic Synoptic Variability (poster). American Geophysical Union Fall Meeting, December 13-17, 2010, San Francisco, CA.

Task Figures

Fig. 1 – MERRA June-July mean sea level pressure for the Arctic Ocean for (a.) 2005-2009.

Fig. 2 – 1979-2000, (b.)

Fig. 3 – (c.) the difference of the two averaging periods, in hPa. Hatching indicates statistical significance at the 95 percent level for the Student’s t.
Posted in Archive

Related Posts

Leave a Reply

Your email address will not be published. Required fields are marked *