Global Precipitation Analysis with the HOAPS-3 Climatology

Axel Andersson¹, Christian Klepp¹, Stephan Bakan²

Statistical approach on brightness temperature (TB) basis:• calibration reference: DMSP F-11• 10-day means of TBs averaged on a global 1°x1° grid, rain-free pixels only•match-ups for each channel and satellite with F-11 for one overlapping year• histogram-equalization of match-ups• linear regression to determine calibration coefficients for TBs of each channel and satellite

• The NAO is the dominating pattern of atmospheric variability in the northern hemisphere• Characterized by the sea-level pressure difference of the Icelandic Low and the Azores high• Analysis based on 1988-2005 winter seasons (Dec-Mar), when NAO is strongest• Station based NAO index (Iceland – Lisbon)(Hurrell, 2003)

H

MILD,

WET

TCOOL,

DRY

COLD

STORMSMIL

D

NAO + phase

T

H

COOL,

DRY

MILD,

WETCOL

D

STORMS

WARM

NAO - phase

HOAPS-3 is combined with the “GPCC Full Data Product Version 4” precipitation data from rain gauge estimates.

This adds over-land observations based on grid interpolated rain-gauge measurements to HOAPS.

The Figure shows the climate mean precipitation of the combined data set for the years 1988-2006.

BUDG

EVAP

WIND

SSTPRECIPCorrelation HOAPS – NAO index

Hamburg Ocean Atmosphere Parameters and Fluxes from Satellite Data

Climate Mean 1988-2005

Global mean time series of satellite and model based precipitation products. Model based estimates exhibit generally higher values. In contrast to satellite retrievals, ERA-40 and NCEP show strong trends

SSM/I intercalibration

Comparison of HOAPS-3 with GPCP V2 (1988-2005) and TRMM 3B43 (1998-2005).

The products show good agreement between 40°N/S. Largest differences occur at the ITCZ and Kuroshio region, where HOAPS-3 exhibits higher values.

In mid and high latitudes GPCP exceeds HOAPS-3 systematically.

HOAPS-3 + GPCC North Atlantic Oscillation (NAO) analysis

EOF Mode 2 of precipitation• Highly correlated to NAO index (r=0.74)

EOF Mode 1 of precipitation (left) • Dominated by event 1989/1990 (abrupt shift from NAO- to strong, persisting NAO+)

www.hoaps.org

Composite patterns of HOAPS-3/GPCC precipitation constructed for the upper/lower quartile of NAO index values. NAO+ pattern shows enhanced storm track and strong precipitation in northern Europe and southeastern US. For NAO- conditions, the storm track is weaker, and heavy precipitation is found in the Mediterranean region.

Combination with GPCC data

Intercomparison with other data sets

Climatological zonal mean for each time series above. Models significantly differ from satellite products in location and magnitude in the tropics. Large differences between all data sets are evident at high latitudes.

Correlation patterns of NAO and HOAPS-3/GPCC precipitation, SST, and over ocean evaporation, wind speed and freshwater flux from HOAPS-3.

Composite NAO+ (upper quartile) Composite NAO- (lower quartile)

The combined HOAPS-3 and GPCC precipitation data set shows realistic and consistent response to the NAO.

¹ University of Hamburg, Germany, ² Max Planck Institute for Meteorology, Hamburg, Germany

HOAPS is a climatology of freshwater flux over the global ocean derived from satellite data.• data availability: 1987-2005 (18 complete years)• 15 parameters: precipitation, evaporation and related surface and atmospheric state parameters• neural net based precipitation algorithm- trained with ECMWF radiative transfer calculations• radiometers on board polar orbiting satellites:- SSM/I (passive microwave)- AVHRR (infrared, SST only)• homogeneous time series: Multi satellite averages containing all SSM/I operating at the same time including inter-sensor calibration• Scan-based, pixel-level dataset (HOAPS-S)• gridded (HOAPS-G, HOAPS-C) datasets , resolution 0.5°, pentad and monthly means, daily composites• data accessible under: www.hoaps.org

Principal components of Mode 1 (dark blue) and NAO index (light blue)

Principal components of Mode 1 (dark blue) and NAO index (light blue)