CRTF Telecon - 05-27-2015

Announcements made prior to the presentations:

- Gil announces that the Reanalysis.org web page will move from its current host (Godaddy.com) to UofColorado/CIRES.  The transition will occur the first week of June, 2015.

- Comments are welcome on the two issues:

1. Reanalysis data formats, and
2. Interest in a shared code base. (will need to login to enter comments into this topic)

Presentations:

 Global Divergence in the new 20^th Century Reanalysis v2

Gil Compo, CIRES-NOAA/OAR/ESRL

Take away points:

1.  Global arguments don’t apply to regional circulations
2.  Observe recent changes in divergence are relatively small compared to variations over the last 100-150 years.

The Walker Circulation is the east-west part of the global overturning circulation in the tropics.  Surface convergence – uplift - upper troposphere divergence areas are over the land masses (South America, Africa, Indonesia) with subsidence in between.

Coupled climate models forced with GHG indicate that global water vapor increases faster than precipitation.  Held and Soden, 2006 say that the global convective mass flux must weaken in order to compensate for this water vapor increase. There is no consensus from other published articles.

Concentrating upon the Pacific Walker Circulation (PWC) and using areas in the western Pacific and eastern Pacific one method used is to take the sea level pressure (SLP) differencing to indicate the strength of the PWC.  The sea surface temperature (Ts) in these two areas can also be differenced.  The SLP and Ts trends and variability are highly correlated and show different trends depending upon which reanalysis or climate model is used.

The PWC is not very well correlated to either the tropical convective mass flux or the global convective mass flux and trends can be of opposite sign, i.e. PWC trends is increasing from 1901 through 2005 while the tropical and global convective mass fluxes show a decreasing trend.

A large scale measure of the divergent circulation is wanted that represents the vertically averaged entire upper atmosphere (above 500 hPa).

Using several reanalyses and SST forced simulations (20CRvr2c, ERA-I, JRA55, and 56 member AMIP using HadlSST 1.1 and and 4 member AMIP using SODAsi SST and COBE-2 sea ice).

The 20CRvr2c data can be accessed at http://go.usa.gov/XTd

A plot of the monthly global anomalies of precipitable water (the integrated column water vapor) from 1851 through 2011 shows that precipitable water decreased during the 1910-1930 period and has been increasing since 1960.  The 20CRv2c and JRA55 precipitable water variability and trends agree well (r=0.75).

Analyses of the vertically average upper half velocity potential for DJF from the 20CRv2c and ERA-I agree very well for the period 2000-2009.  The gradient of the velocity potential contours gives the vertically average upper tropospheric divergent wind. Comparing the 2000-2009 period versus the 1980-1999 period shows that the divergence pattern has been expanding westward and also poleward (Hadley circulation). Comparison of longer periods (1960-2009) and (1910-1959) shows very little difference. 

To evaluate the long term trend of the divergence, an index for global divergent circulation is determined by using the DJF minimum value and applying an 11 year smoother to reduce inter-annual variability. The 20CRv2c shows a stronger global divergent circulation than the original 20CR from 1850-1940 with less variability.  They both show weakened divergent circulation from the 1950’s though 1990 and strengthening since 1990 through the present.  The JRA55 agrees well with the 20CRv2c over its 1958-present and ERA-I agrees well over its 1979-present periods.  Both reanalyses show the strengthening divergent circulation since 1990.

Conclusions:

1. Pacific Walker Circulation trends and variability depend upon how they are defined.  A SLP-based definition is almost unrelated to global and tropical overturning circulation.
2. SLP-based PWC index is not a proxy for the global or tropical convective mass flux. i.e. global arguments cannot be applied to regional circulation.
3. Broaden scope to examine Global Divergent Circulation using large-scale vertically-integrated upper-level divergence.
4. Recent strengthening of DJF Global Divergent Circulation is small compared to variations over the past 150 years.

Quantifying the uncertainties of reanalyzed Arctic cloud-radiation properties using satellite-surface observations

Yiyi Huang and Xiquan Dong, U of North Dakota

Data Set Info:

• Time frame: 2000-2012
• Area: Arctic latitudes (70-90N)
• Satellite obs: CERES-MODIS cloud fraction (CF) and CERES-EBAF (Energy Balanced and Filled) for TOA and surface fluxes (see https://climatedataguide.ucar.edu/climate-data/ceres-ebaf-clouds-and-earths-radiant-energy-systems-ceres-energy-balanced-and-filled for more information)
• Reanalyses: JRA55, 20CRv2c, CFSR, ERA-I, and MERRA

Methodology:

• Generate monthly means of cloud and radiation parameters
• Data averaged over 2x2 degree grid boxes.

Comparison between satellite-derived and surface BSRN observations.

Comparison of CERES CF and reanalyses shows only JRA55 captures the seasonal variability and spatial distribution of CF however it is biased lower than CERES CF.  The other reanalyses do not show much seasonal variability.  The CERES CF may have greater uncertainty during Arctic winter months.

TOA LW up radiative fluxes: All reanalyses capture seasonal cycle of LWup All Sky and Clear Sky values. 20CRv2c has negative bias in winter months for All Sky implying that its CF is too great during the winter.  JRA55 has a positive bias all year for All Sky implying that its CF is too small compared to CERES CF.  There is good agreement between all the reanalyses and CERES for Clear Sky.  The reanalyses have a slight positive bias.

SWup: All Sky peaks in June while Clear Sky peaks in May due to higher solar radiation and snow and ice coverage.  Reanalyses agree with seasonal cycle but have variability of peak amounts in warm season.

Cloud Radiative Effect (CRE): LW CREs are positive indicating a warming effect. Reanalyses vary on seasonal amount. TOA SW are negative with greatest cooling effect in JJA. Reanalyses agree with seasonal cycle but vary on minimum amount in July. Net CREs are dominated by LW warming during winter months and SW cooling during summer.  Good agreement between the reanalyses on the net CRE.

Warm season (JJA) TOA net CREs: CERES shows strongest warming effect over Greenland associated with low CF.  Strongest cooling effect occurs over the North Atlantic Ocean with higher CF.

Surface SW radiation: SWup is greatest in May.  MERRA is much lower.  SWdn is greatest in June.  MERRA is lower.  Other reanalyses agree more with seasonal cycle with variability about the peak values.  Spatially, CERES-EBAF has max SWdn over Greenland where CF is low.  Minimum values are over North Atlantic Ocean were CF is higher.  CFSR, ERA-I and MERRA are negatively biased in the Arctic Ocean. CFSR and ERA-I SWdn are consistent with CF results.

Summary:

1. Except JRA55 the other four reanalyses have difficulties in representing the seasonal variation of Arctic cloud faction, especially during the cold season where they overestimate.
2. Almost all the reanalyses can well capture the seasonal cycle of radiation fluxes, both at the TOA and Surface.
3. TOA net CREs are dominated by LW warming effect during winter months and SW cooling effect from April to September.
4. In general, most of the reanalyzed CF and radiation results are not physically consistent with each other, such as higher C should result in higher TOA SWup and lower Surface SWdn flux.

-Annarita pointed out the Arctic System Reanalysis can be used for comparison.

-It was pointed out that the uncertainty of the satellite and surface radiation observations may overlap such that their differences may not be that great.