ATMS Snowfall Rate Product and Its Applications

Huan Meng1, Ralph Ferraro1, Cezar Kongoli2, Jun Dong2, Bradley Zavodsky3,

Banghua Yan1, Limin Zhao1, Nai-Yu Wang2

1NOAA/National Environmental Satellite, Data, and Information Service

2University of Maryland/ESSIC/Cooperative Institute for Climate and Satellites

3NASA/Short-term Prediction Research and Transition Center

Introduction

• The ATMS Snowfall Rate (SFR) product is water equivalent snowfall rate estimate over

global land Cross-track scanning with mixed polarizations

Channels from 23 GHz – 183 GHz

• The algorithm partially inherits the operational AMSU/MHS SFR, but with many new

developments that lead to superior

performance

• Currently, SFR is generated from five satellites (S-NPP and four POES and Metop

satellites) with about ten estimates per day in

mid-latitudes and more in high latitudes

• The ATMS SFR algorithm was developed with the support of JPSS Proving Ground and Risk

Reduction program

Retrieved Snowfall Rate

Composite NEXRAD Reflectivity

2

3

• SFR algorithm includes two main components

Snowfall detection – statistical algorithm

Snowfall rate – physically-base algorithm

Methodology

Snowfall Detection Algorithm

Satellite-based module

Coupled principal component and logistic

regression model

Use all seven high-frequency channels at and

above 88.2 GHz and the temperature sounding

channel at 53.6 GHz

Two temperature regimes

Two cloud thickness regimes

Trained with gauge observations

NWP model-based module

Logistic regression model

Optimal combination of the two modules

Output is probability of snowfall; use preset

thresholds to determine snowfall

Additional NWP model-based screenings

Improves the accuracy of snowfall detection

4

Snowfall

Probability

Snowfall

Rate

Snowfall

Detection

Index

The combined SD algorithm improves detection of both shallow-

and thick-cloud snowfalls

Combined SD Algorithm

Single-Module SD Combined SD Radar Reflectivity

Shallow-Cloud

Snowfall Case

Thick-Cloud

Snowfall Case

6

• 1DVAR retrieval of cloud properties Forward simulation of Tb’s (5 frequencies, window and water

vapor sounding channels) with a radiative transfer model

(RTM) (Yan et al., 2008)

Iteration scheme with predefined ΔTb thresholds

IWP and De are retrieved when iteration stops

• Computation of ice particle fall velocity Heymsfield and Westbrook (2010)

• Determination of snowfall rate Adjusting factor to compensate for non-uniform ice water

content distribution in cloud column; derived from collocated

satellite and StageIV radar and gauge combined hourly

precipitation data

Snowfall Rate Algorithm

SFR Recalibration & Validation

7

Recalibration using Multi-Radar Multi-Sensor (MRMS) instantaneous

snowfall rate data to reduce a dry bias - histogram matching (Kidder and

Jones, 2007) to adjust SFR CDF towards MRMS

Validation against MRMS: Six multi-day snowfall events, 7794 matching

data points

Correlation

Coefficient

Bias

(mm/hr)

RMS

(mm/hr)

Original 0.55 -0.30 0.77

Recalibrated 0.56 -0.10 0.73

Correlation

Coefficient

Bias

(mm/hr)

RMS

(mm/hr)

0.52 -0.07 0.75

• Most blended satellite precipitation datasets do not include satellite snowfall

rate product – use other data sources

(model, ground observations, etc.)

• CMORPH is a NOAA global blended precipitation analysis product with wide-

ranging applications

• The first generation CMORPH only has rain rate. The SFR product is integrated

in the second generation CMORPH

• A sample for a major snowstorm (right) Stage IV radar precipitation image (bottom)

shows a warm band (rainfall) and a cold band

(snowfall) of precipitation from a frontal system

The second generation CMORPH (top)

captures both bands after integrating SFR

• National Water Center uses SFR through the use of CMORPH

Application in Hydrology

(Xie and Joyce, NOAA/NCEP/CPC)

Stage IV

Radar Precip

2nd Generation

CMORPH

Snowfall

Rainfall

8

Application in Weather Forecasting

SFR assessment at several National Weather Service Weather Forecast Offices in

a NASA SPoRT project. User feedback

indicates that SFR is a useful product for

weather forecasting operations

SFR assessment at the Weather Prediction

Center Hydrometeorological Testbed’s Winter

Weather Experiment (WWE) this winter.

SFR is especially useful for filling

observational gaps in mountains and remote

regions where radar and weather stations are

sparse or radar blockage and overshooting

are common

SFR also provides quantitative snowfall

information to complement snowfall

observations or estimations from other

sources

A radar and SFR combined product, mSFR,

with 10-min interval

9

Radar Precip Quality Index

during 2016 East Coast Blizzard

poor coverage

quality degradation

during snowfall

WWE probabilistic snowfall rate

forecast overlaid on SFR

SFR Using Direct Broadcast Data

Reduce latency to meet requirement for weather

forecasting – forecasters’ feedback

Retrieve DB CONUS and Alaska L1B data from Univ. of

Wisconsin, Madison/CIMSS

Generate SFR within 30 min of observation; SFR with

operational L1B data has 30 min ~ 3 hr delay

Output:

Data made available to NASA/SPoRT, reformat to AWIPS

II/NAWIPS, and disseminate to WFOs and WPC

Images posted on SFR webpage at near real-time

Webpage:

NESDIS/CICS:

http://cics.umd.edu/sfr

http://www.star.nesdis.noaa.gov/corp/scsb/mspps_backup/sfr_realtim

e.html

SPoRT:

http://weather.msfc.nasa.gov/cgi-

bin/sportPublishData.pl?dataset=snowfallrateconus&product=conus_s

nowrate

10

Case 1, December 14, 2014

Albuquerque, NM WFO (ABQ): The product (SFR) did validate that we will be able to complement radar void coverage areas in an operational forecast environment using polar-orbiting satellite imagery.

Ground

reported

snowfall

11

• The 2016 Blizzard hit the Mid-Atlantic region on 22-24 January 2016 and produced record snowfall in many local areas

• The ATMS and MHS SFR products captured the evolution of the blizzard with five satellites including S-NPP, POES and Metop.

Correl. Coeff.

Bias (mm/hr)

RMS (mm/hr)

ATMS 0.60 -0.14 0.79

MHS 0.54 -0.53 0.88

Case 2, January 22-24, 2016

Jan 23 07:13Z S-NPP

SFR

MRMS

Jan 23 18:39Z S-NPP

12

SFR SFR

Blizzard of 2016 (By Patrick Meyers, CICS-MD)

14

mSFR – Radar and Satellite Merged SFR

Summary

An ATMS Snowfall Rate product has been developed with the JPSS

PGRR support

Extensive validation studies have demonstrated the quality of the product

Product applications to support NWS

Hydrology: CMORPH (CPC), NWC

Weather Forecasting: WFOs, WPC

SFR is generated within 30 min at STAR/CICS using direct broadcast

data

A radar-satellite fused snowfall rate product has been developed and

generated at real-time

Future Plan Development of SSMIS SFR algorithm

Development of GMI SFR algorithm

Development of prototype ocean SFR algorithms

15