DoD Center for Geosciences/Atmospheric Research at Colorado State University VTC 12 September 2011 1 Global Precipitation Products for Data-Denied Regions.

DoD Center for Geosciences/Atmospheric Research at Colorado State University VTC 12 September 20111

Global Precipitation Products for Data-Denied Regions

Stanley Q. KidderCenter for Geosciences/Atmospheric Research

Colorado State University

DoD Center for Geosciences/Atmospheric Research at Colorado State University VTC 12 September 20112

Outline

Review of current satellite rainfall methods (Add sites to be covered)

Outline of CIRA products Blended RR Blended TPW ORI

Critique of current RR methods/outline of possible techniques Need motion Need accumulation Need orographic enhancement Need TPW for forecasts

The proposal

DoD Center for Geosciences/Atmospheric Research at Colorado State University VTC 12 September 20113

Importance

Detection of precipitation is critical for the success of military operations. Air operations at sea or over land, mobilization of heavy equipment on third world roads, operation of laser-based systems, and deployment of resources in flood-prone regions all rely on timely and accurate prediction of rainfall events.

DoD Center for Geosciences/Atmospheric Research at Colorado State University VTC 12 September 20114

History

Because of its importance, precipitation has been estimated from satellite observation since the earliest weather satellites

Two (of many) sources of information about these early attempts: Barrett, E. C., and D. W. Martin, 1981: The Use of Satellite Data

in Rainfall Monitoring. Academic Press, New York. Kidder, S. Q., and T. H. Vonder Haar, 1995: Satellite Meteorology:

An Introduction. Academic Press, San Diego.

DoD Center for Geosciences/Atmospheric Research at Colorado State University VTC 12 September 20115

Current “Real Time” Precipitation Sources

The University of California Irvine PERSIANN project (Sorooshian et al. 2000) (See http://chrs.web.uci.edu/persiann/)

uses GEO IR and visible imagery in a neural network system to produce global precipitation grids. The PERSIANN system has since been extended to use TRMM, NOAA and DMSP microwave data to “calibrate” higher resolution IR data.

DoD Center for Geosciences/Atmospheric Research at Colorado State University VTC 12 September 20116

Current “Real Time” Precipitation Sources

The Climate Prediction Center’s CMORPH/QMORPH (Joyce et al. 2004) (See http://www.cpc.ncep.noaa.gov/products/janowiak/cmorph_description.html)

Based on microwave observations calibrated with TRMM data. Uses IR winds to move raining pixels between satellite obervations. QMORPH is the forecaster product, and moves precipitating pixels forward in time to the hour or half hour.

DoD Center for Geosciences/Atmospheric Research at Colorado State University VTC 12 September 20117

Current “Real Time” Precipitation Sources

CIRA’s Blended Rain Rate (http://cat.cira.colostate.edu) Being developed for operational implementation at NESDIS

Blends data from currently six satellites (see next slide) Uses a histogram-matching technique to make it appear that all observations are from the same

instrument No motion.

DoD Center for Geosciences/Atmospheric Research at Colorado State University VTC 12 September 20118

Current Blended RR Satellite Suite

DoD Center for Geosciences/Atmospheric Research at Colorado State University VTC 12 September 20119

Regularly Spaced Constellation

DoD Center for Geosciences/Atmospheric Research at Colorado State University VTC 12 September 201110

Current “Real Time” Precipitation Sources

NASA Goddard’s TRMM Multi-satellite Precipitation Analysis (TMPA) (Huffman et al. 2007) (See http://precip.gsfc.nasa.gov/index.html)

Combines TMI , AMSR-E, and AMSU-B/MHS microwave measurements. Gaps are filled using IR data. No motion Latency is an issue (~5 hours)

DoD Center for Geosciences/Atmospheric Research at Colorado State University VTC 12 September 201111

Current “Real Time” Precipitation Sources

The NESDIS Hydro-Estimator (Scofield and Kuligowski, 2003) (See http://www.star.nesdis.noaa.gov/smcd/emb/ff/HydroEst.php)

Uses GEO IR data in a single channel algorithm GFS model low level humidity estimates to adjust for sub-cloud evaporation over arid regions. No microwave data are used.

DoD Center for Geosciences/Atmospheric Research at Colorado State University VTC 12 September 201112

Current “Real Time” Precipitation Sources

The Naval Research Laboratory (NRL) Blended-Satellite Precipitation Technique (Turk and Miller 2005) (See http://www.nrlmry.navy.mil/sat-bin/rain.cgi)

uses GEO IR observations, calibrated using passive microwave observations, plus the microwave observations themselves (when available) to produce 3-hr rainfall accumulations.

DoD Center for Geosciences/Atmospheric Research at Colorado State University VTC 12 September 201113

Summary of the Current State

IR provides frequent, high-resolution observations but IR looks only at the outside of the cloud; precipitation must be inferred.

Microwaves view the precipitation-size droplets or crystals inside the cloud, but at lower frequency and resolution.

Current products are mostly aimed at instantaneous rainfall rates, whereas forecasters need accumulated rainfall

To calculate accumulation—given the motion of raining systems—some sort of propagation scheme must be employed

DoD Center for Geosciences/Atmospheric Research at Colorado State University VTC 12 September 201114

QMORPH 5-Day (6-hour)

DoD Center for Geosciences/Atmospheric Research at Colorado State University VTC 12 September 201115

QMORPH 1-Day (hourly)

DoD Center for Geosciences/Atmospheric Research at Colorado State University VTC 12 September 201116

Blended RR 5-Day (6-hourly)

DoD Center for Geosciences/Atmospheric Research at Colorado State University VTC 12 September 201117

Blended RR 1-Day (hourly)

DoD Center for Geosciences/Atmospheric Research at Colorado State University VTC 12 September 201118

5-Day Comparison

DoD Center for Geosciences/Atmospheric Research at Colorado State University VTC 12 September 201119

1-Day Comparison

DoD Center for Geosciences/Atmospheric Research at Colorado State University VTC 12 September 201120

From Kelly Howell’s Thesis

y = 0.0224x1.8551

R² = 0.9985

0%

5%

10%

15%

20%

25%

30%

35%

40%

45%

50%

0 - 15 mm 15 - 30 mm 30 - 45 mm 45 - 60 mm 60 - 75 mm

Prob

. of P

reci

p.

TPW Range

Threshold = 0.1 mm hr-1

TPW might be useful for

determining how long precipitation

lasts

DoD Center for Geosciences/Atmospheric Research at Colorado State University VTC 12 September 201121

What Is Needed

3 hr accumulations Short-term forecast (nearcast) Motion Orographic enhancement

DoD Center for Geosciences/Atmospheric Research at Colorado State University VTC 12 September 201122

Blended Total Precipitable Water

DoD Center for Geosciences/Atmospheric Research at Colorado State University VTC 12 September 201123

06Z 14 Oct 2009

250 50 75 mm

Blended TPW (mm) and 850 hPa GFS Winds

DoD Center for Geosciences/Atmospheric Research at Colorado State University VTC 12 September 201124

Elevation (m)

0 1000 2000 3000 4000

Terrain

DoD Center for Geosciences/Atmospheric Research at Colorado State University VTC 12 September 201125

ORI TPW * V•H

Orographic Rain Index

TPW = Blended Total Precipitable Water advected to product time

V = 850 hPa GFS wind vector at product time

H = Terrain height

V•H terrain-induced vertical velocity

ORI moisture * liftUnits = mm * m/s

DoD Center for Geosciences/Atmospheric Research at Colorado State University VTC 12 September 201126

Observed precipitation (inches) 06Z Tue 13 Oct 15Z Wed 14 Oct

00Z 14 Oct 2009

250 +0 50 100 150 200

DoD Center for Geosciences/Atmospheric Research at Colorado State University VTC 12 September 201127

Observed precipitation (inches) 06Z Tue 13 Oct 15Z Wed 14 Oct

00Z 14 Oct 2009

DoD Center for Geosciences/Atmospheric Research at Colorado State University VTC 12 September 201128

Global Precipitation Products for Data-Denied Regions

We propose to develop a blended rain accumulation product which addresses some of these problems. In outline, the product might be constructed as follows:

1. The instantaneous satellite rain rate observations, both infrared and microwave, would be gathered for an integration period of perhaps three hours. (Though at first only microwave observations would be used.)

2. The instantaneous observations would be “calibrated” using histogram adjustment to a reference instrument, so that artifacts due to differing instruments would be lessened.

3. The rain observations would be propagated in time using model winds, and perhaps TPW values, which could possibly be used to estimate changes in intensity of the observed rain rates in time.

4. Rain accumulations during the analysis period would be calculated for each grid box of the analysis domain.

5. Terrain induced rain would be added to the accumulations.

6. Model forecasted rain accumulations could be used to modify/constrain the satellite-estimated accumulations.