Contrasting Tropical Rainfall Regimes Using TRMM and Ground-Based Polarimetric Radar

by

S. A. Rutledge, R. Cifelli, T. Lang and S. W. Nesbitt

EGU 2009

Session AS1.3

•The TRMM satellite has provided unprecedented data for over 10 years. TRMM precipitation products have advanced our understanding of tropical precipitation considerably.

•We used ground-based polarimetric radar data from two field campaigns, TRMM-LBA (1999) and NAME (North American Monsoon Experiment; 2004), to characterize the physical nature of precipitation in these regions (tropical Brazil and Mexico).

•We focus on comparing rain rate and reflectivity distributions between the TRMM PR and the ground based radar data.

TRMMLBA

NAME

East-West Regimes found in TRMM-LBA.East regime associated with high CAPE, vigorous convection and frequent lightning.West regime associated with lower CAPE,weaker convection and lower flash rates. January-February 1999.

•Strong topographical influence on precipitation intensity and structure.

For both NAME and LBA, observationssuggest equilibrium DSD indicatedby near constant D0 values at high rain rates/reflectivities.

Mean TRMM reflectivity values are lower compared to S-pol below 7-8 km.Above this level TRMM PR is higher mainly due to 17 dB cut off for the PR.

Are the reflectivity differences below 7-8 km a result of attenuation correction issues for the PR (wet ice, Mie effects), sampling issues (many more PR profiles compared to ground based profiles), or other factors?

S-pol resolution has been degraded

For NAME, PR reflectivities are again less than those for S-pol in a vertical profilesense. Near surface values are in better agreement with S-pol.

In a mean profile sense, and statistically, the PR reflectivities are lower compared to ground based observations for both locales.

PR profile remains belowS-pol profile.Stronger convection inNAME.

TRMM PR does not see the high rainrates observed in the S-pol spectrum

TRMM PR places more rain volumeat lower rain rates compared to theground based radar.

Turn now to examination of rain rates between PR and S-pol for TRMM-LBA

NAME shows similar behavior, especially over water and the coastal plains.At higher elevations, TRMM PR and S-pol rain rate spectrum is in better agreement. This may be due to reduced S-pol resolution (compared to PR) at long ranges required to sample storms over the higher terrain.

Over water Coastal plain

500-1500 m > 1500 m

And the sameanalysis for NAME

For NAME, TRMM PR places more rain volume at lower rain rates compared to theground based radar. Consistent with finding for TRMM-LBA.

10-15% differences in rain volume seen for this particular comparison.

Z-R plot for LBASolid--S-polDash--PR

Above 10 mm/h,PR reflectivitycorresponds tolower rain ratecompared to samereflectivity for S-pol.

Over water Over coastal plains

Z-R relationships in fair agreement over water.

Z-R relationships in lesser agreement over adjacent coastal plains and higher topographical elevations (not shown).

NAME

For a given Z , TRMM Z-R producesless rain compared to S–pol for R > 10 mm/h

Summary• To the extent that the differences in the PR and ground based data records don’t

matter…..• TRMM PR reflectivities reduced over land compared to ground based observations

in two field projects. Does this reflect issues with attenuation correction over land?

• TRMM generated rain rates are lower than ground based radar estimates for a given reflectivity.

• Single polarization, radar-based rain estimates such as those from the PR are not capable of capturing the full physics of the DSD. So even if the Z’s were to match….

• Evidence for equilibrium DSD being achieved in these tropical rain regimes; drop break up produces smaller drops. This process complicates Z based rain estimation.

• Appeal next to Darwin dataset that is much longer than field campaign data

Acknowledge long term support from NASA TRMM and PMM programs

DSD differences can beseen in distribution of Acoefficient in Z = ARb

DSD

Attenuation correction

Rain estimation