combining glm and abi data for enhanced goes-r rainfall estimates

TEMPLATE DESIGN © 2008 www.PosterPresentations.com Combining GLM and ABI Data for Enhanced GOES-R Rainfall Estimates Robert F. Adler, Weixin Xu, and Nai-Yu Wang CICS, University of Maryland, College Park, MD. Email: [email protected] This study uses TRMM data to develop the basis and eventually the algorithms to combine GOES-R ABI Infrared (IR), Geostationary Lightning Mapper (GLM) data to provide an improved geosynchronous rainfall product. A simple IR- lightning rainfall algorithm is developed through coupling lightning measurements into the Convective/Stratiform (C/S) technique (CST) proposed by Adler and Negri (1988). Preliminary results show that CST is improved by more than 30% in retrieving convective (or heavy) rainfall after lightning observations are coupled. Specifically, lightning information can aid in identifying convective cores missed by the IR-only technique (improving detection), eliminate misidentified convective cores (lowering false alarms), and more correctly assign heavy convective rainfall rates. Combining Infrared and Lightning Information for Rainfall Estimates (Applied to TRMM data) 2. IR-based C/S Technique (CST) 3. Lightning-enhanced CST (CSTL) 4. Evaluation on CST and CSTL Estimates B. Estimates on Instantaneous Rainrate C. Comparison with SCaMPR (GOES) 1. Introduction a. Conv. cores w/o lightning in mature systems are removed; b. Conv. areas (with flashes) missed by CST are added; c. Rain rates are derived from both IR and lightning density; CST algorithm tuned from Adler and Negri (1988) a. Conv. core defined by local Tb minima that pass slope test; b. Conv. core area is an empirical function of Tb minima. CST ( and most IR techniques) does a GOOD job in catching young convective cells. CST ( and most IR techniques) becomes VAGUE as mesoscale convective systems mature (too many convective cores). Lightning is very valuable in helping IR technique to correctly identify heavy rainfall. PWM RR Radar RR IR+L RR SCaMPR RR CST C/S Radar C/S TMI C/S (10mm/hr) CST C/S IR Tb IR Tb + lightning CST RR CST + L RR PR RR TMI RR Dependent data (used to train the algorithm): 2002-2004, May-August Independent data: 2005-2008, May-August Both datasets include only storms > 5 flashes per minute A. Estimates on Convective Rain 20 km res. May 2007 100 cases (> 5 flash/min) CST PR CST TMI CSTL PR TMI POD FAR CSI CSTL Correlation Coefficient (CC) BIAS & RMSE 20 km res. May-Aug 2002- 2008 3500 cases (> 5 flash/min)

Upload: latika

Post on 25-Feb-2016

27 views

Category:

Documents

0 download

Report

Download

Embed Size (px):

DESCRIPTION

Combining GLM and ABI Data for Enhanced GOES-R Rainfall Estimates Robert F. Adler, Weixin Xu, and Nai-Yu Wang CICS, University of Maryland, College Park, MD. Email: [email protected]. Combining Infrared and Lightning Information for Rainfall Estimates (Applied to TRMM data) . - PowerPoint PPT Presentation

TRANSCRIPT

Page 1: Combining GLM and ABI Data for Enhanced GOES-R Rainfall Estimates

www.PosterPresentations.com

Combining GLM and ABI Data for Enhanced GOES-R Rainfall EstimatesRobert F. Adler, Weixin Xu, and Nai-Yu Wang

CICS, University of Maryland, College Park, MD. Email: [email protected]

This study uses TRMM data to develop the basis and eventually the

algorithms to combine GOES-R ABI Infrared (IR), Geostationary

Lightning Mapper (GLM) data to provide an improved geosynchronous

rainfall product. A simple IR-lightning rainfall algorithm is developed

through coupling lightning measurements into the Convective/Stratiform

(C/S) technique (CST) proposed by Adler and Negri (1988). Preliminary results show that CST is improved by more than 30% in retrieving convective (or heavy) rainfall after lightning observations are coupled. Specifically, lightning information can aid in identifying convective cores missed by the IR-only technique (improving detection), eliminate misidentified convective cores (lowering false alarms), and more correctly assign heavy convective rainfall rates.

Combining Infrared and Lightning Information for Rainfall Estimates (Applied to TRMM data)

2. IR-based C/S Technique (CST)

3. Lightning-enhanced CST (CSTL)

4. Evaluation on CST and CSTL Estimates

B. Estimates on Instantaneous Rainrate

C. Comparison with SCaMPR (GOES)

1. Introduction

a. Conv. cores w/o lightning in mature systems are removed; b. Conv. areas (with flashes) missed by CST are added;c. Rain rates are derived from both IR and lightning density;

CST algorithm tuned from Adler and Negri (1988)a. Conv. core defined by local Tb minima that pass slope test;b. Conv. core area is an empirical function of Tb minima.

CST ( and most IR techniques) does a GOOD job in catching young convective cells.

CST ( and most IR techniques) becomes VAGUE as mesoscale convective systems mature (too many convective cores).

Lightning is very

valuable in helping

IR technique to

correctly identify

heavy rainfall.

PWM RR Radar RR

IR+L RRSCaMPR RR

CST C/SRadar C/S

TMI C/S (10mm/hr) CST C/S

IR Tb

IR Tb + lightning

CST RR CST + L RR

PR RRTMI RR

Dependent data (used to train the algorithm):2002-2004, May-AugustIndependent data: 2005-2008, May-AugustBoth datasets include only storms > 5 flashes per minute

A. Estimates on Convective Rain

20 km res.

May 2007

100 cases(> 5 flash/min)

CST

TMI

CSTL

TMI

POD

FAR CSI

CSTL

Correlation Coefficient (CC)