AIRS Data Assimilation at the Regional Scale Brad Zavodsky & - - PowerPoint PPT Presentation

▶
airs data assimilation at the regional scale
SMART_READER_LITE
LIVE PREVIEW

AIRS Data Assimilation at the Regional Scale Brad Zavodsky & - - PowerPoint PPT Presentation

Apr 11, 2024 527 likes •696 views

Earth Science Office Science Mission Directorate UAH UAH UAH National Aeronautics and Space Administration National Aeronautics and Space Administration AIRS Data Assimilation at the Regional Scale Brad Zavodsky & Will McCarty

slide-1
SLIDE 1

Science Mission Directorate

National Aeronautics and Space Administration

transitioning unique NASA data and research technologies to the NWS

UAH UAH UAH

AIRS Data Assimilation at the Regional Scale

Brad Zavodsky & Will McCarty

University of Alabama in Huntsville Huntsville, Alabama

Shih-hung Chou, Gary Jedlovec, & Bill Lapenta

NASA / Marshall Space Flight Center Huntsville, Alabama

AIRS Science Team Meeting – March 2007

Earth Science Office

National Aeronautics and Space Administration

slide-2
SLIDE 2

Science Mission Directorate

National Aeronautics and Space Administration

transitioning unique NASA data and research technologies to the NWS

UAH UAH UAH

Outline

  • SPoRT AIRS assimilation focuses on short-term

regional forecasts—compliments work at JCSDA

  • Profile Assimilation
  • Lessons Learned from Case Study
  • Analysis Impact (v4.13; GSFC)
  • v4/v5 Profile Comparison w/ Rawinsonde (v4.0, v5.0; JPL)
  • Near-Real-Time (NRT) Assimilation plans
  • Radiance Assimilation
  • Determination of Uncontaminated Channels
  • Validation with MODIS/CloudSat

Earth Science Office

National Aeronautics and Space Administration

slide-3
SLIDE 3

Science Mission Directorate

National Aeronautics and Space Administration

transitioning unique NASA data and research technologies to the NWS

UAH UAH UAH

Motivation for Profile Assimilation at SPoRT

  • AIRS profiles complement traditional upper-air observations

in data sparse regions (e.g. ocean).

  • Hyperspectral nature of AIRS sounder allows for highest

vertical resolution of any current remote sensing system

  • L2 profiles provide a data set to add information to initialize

forecast models in data-void regions without running complex RTA within analysis

  • What follows is an overview of work done with v4, some

preliminary work with v5, and an overview of upcoming real- time work

Earth Science Office

National Aeronautics and Space Administration

slide-4
SLIDE 4

Science Mission Directorate

National Aeronautics and Space Administration

transitioning unique NASA data and research technologies to the NWS

UAH UAH UAH

Insights from Previous Case Study Work

  • Short WRF forecast initialized with NAM used as background for ADAS

analysis; 48-hour WRF forecasts for Nov. 2005 east coast storm

  • AIRS profiles (v4.13) have a positive impact on the initial conditions of

the model (next slide) but have varying results on regional forecasts with improvements at some forecast times at some levels

  • AIRS impact on forecast depends on case study, use of QIs, assimilation

time (model adjustment), and model resolution

Surface analysis 11/20/05 12 UTC Surface analysis 11/20/05 12 UTC

L

Surface analysis 11/21/05 12 UTC Surface analysis 11/21/05 12 UTC

L

Surface analysis 11/22/05 12 UTC Surface analysis 11/22/05 12 UTC

L

L L L WRF Domain for Nov. 2005 Case Study WRF Domain for Nov. 2005 Case Study

Earth Science Office

National Aeronautics and Space Administration

slide-5
SLIDE 5

Science Mission Directorate

National Aeronautics and Space Administration

transitioning unique NASA data and research technologies to the NWS

UAH UAH UAH

AIRS V5 PBest

Initial Assessment of V5 Profiles: Sept. 8, 2006

GOES-11 IR Image AIRS V4 RetQAFlag from DAAC

  • Substantially more full, high quality retrievals over land (Midwest)
  • Data removed mainly in cloudy regions
  • V5 quality control adds data over land, near clouds, and above clouds

AIRS V5 PBest from JPL Focus Day

Earth Science Office

National Aeronautics and Space Administration

slide-6
SLIDE 6

Science Mission Directorate

National Aeronautics and Space Administration

transitioning unique NASA data and research technologies to the NWS

UAH UAH UAH

20 November 2005 MHX 07Z BKGD 07Z AIRS 07Z ADAS 00Z RAOB 12Z RAOB 07Z BKGD 07Z AIRS 07Z ADAS 00Z RAOB 12Z RAOB 20 November 2005 WAL

Impact of AIRS Profiles on Initial Conditions

AIRS detects lowering of moist layer (00-12Z) AIRS correctly detects 850 – 700 hPa dry layer AIRS correctly shows mid- troposphere cooling (00-12Z) AIRS shows moistening with time (00-12Z)

Earth Science Office

National Aeronautics and Space Administration

slide-7
SLIDE 7

Science Mission Directorate

National Aeronautics and Space Administration

transitioning unique NASA data and research technologies to the NWS

UAH UAH UAH

Detroit, MI (DTX): 5 km Greensboro, NC (GSO): 0 km Land Soundings (LandFrac ≥ 0.50)

00Z RAOB 12Z RAOB V4 AIRS V5 AIRS 00Z RAOB 12Z RAOB V4 AIRS V5 AIRS

Too dry below 800 hPa for both V4 and V5; slight degradation in V5 T appears more consistent with RAOBs for v5 throughout atmosphere

Earth Science Office

National Aeronautics and Space Administration

q doesn’t perfectly depict moisture for all cases

slide-8
SLIDE 8

Science Mission Directorate

National Aeronautics and Space Administration

transitioning unique NASA data and research technologies to the NWS

UAH UAH UAH

Key West, FL (EYW): 69 km

Wallops Island, VA (WAL): 58 km

Water Soundings (LandFrac = 0.0)

00Z RAOB 12Z RAOB V4 AIRS V5 AIRS 00Z RAOB 12Z RAOB V4 AIRS V5 AIRS

Smaller change between v4 and v5 in T and q (mostly positive compared to RAOBs)

Earth Science Office

National Aeronautics and Space Administration

slide-9
SLIDE 9

Science Mission Directorate

National Aeronautics and Space Administration

transitioning unique NASA data and research technologies to the NWS

UAH UAH UAH

  • Single case studies are not necessarily representative (statistically significant) of
  • verall model performance
  • Looking to test sensitivity and feasibility (e.g. make future forecasts or initial

conditions available to WFOs) of AIRS data in real time; not trying to run optimal

  • perational configuration
  • CNTL: control; use no AIRS data
  • AIRS: use QIs and error profile information to select only the highest quality data
  • Use real time assimilation to select focus days for further study

Real-Time Assimilation

L

Looking for cases with surface feature like this

Earth Science Office

National Aeronautics and Space Administration

0600 0900 1200

Desired initializatio n time NAM data available WRF bkgd ready for 0700 ADAS 1st AIRS available 1st AIRS valid 2nd AIRS valid 0700 analysis complete WRF bkgd ready for 0900 ADAS 2nd AIRS available* 0900 analysis complete WRF forecast complete as next NAM forecast cycle begins

slide-10
SLIDE 10

Science Mission Directorate

National Aeronautics and Space Administration

transitioning unique NASA data and research technologies to the NWS

UAH UAH UAH

Outline

  • SPoRT AIRS assimilation focuses on short-term

regional forecasts—compliments work at JCSDA

  • Profile Assimilation
  • Lessons Learned from Case Study
  • Analysis Impact (v4.13; GSFC)
  • v4/v5 Profile Comparison w/ Rawinsonde (v4.0, v5.0; JPL)
  • Near-Real-Time (NRT) Assimilation plans
  • Radiance Assimilation
  • Determination of Uncontaminated Channels
  • Validation with MODIS/CloudSat

Earth Science Office

National Aeronautics and Space Administration

slide-11
SLIDE 11

Science Mission Directorate

National Aeronautics and Space Administration

transitioning unique NASA data and research technologies to the NWS

UAH UAH UAH

Motivation for Radiance Assimilation at SPoRT

  • Like profiles, radiances can be used to

supplement rawinsondes in data sparse regions

  • Traditional cloud detection approaches may be

too conservative for mesoscale variability important for regional assimilation studies

  • Additional cloud-free channels may add

mesoscale detail

  • Enhanced CO2 sorting technique is applied to

AIRS radiances to this end (Will McCarty, JCSDA)

  • What follows is a brief description of this

technique and some validation against MODIS and CloudSat observations

Earth Science Office

National Aeronautics and Space Administration

Technique used for IFOV2

slide-12
SLIDE 12

Science Mission Directorate

National Aeronautics and Space Administration

transitioning unique NASA data and research technologies to the NWS

UAH UAH UAH

Determination of Usable Channels in IFOV2

clear IFOV cloudy IFOV

separation point

  • CO2 sorting technique (Holz et al. 2006) adapted to

distinguish between contaminated and uncontaminated radiances

  • Clear spectrum generated using forward RT

calculation; sorted with cloudy spectra by BT to determine separation point between clear and cloudy channels

  • SPoRT sorting

technique (left figure) compares well to the CO2 slicing CTP (right figure)

Earth Science Office

National Aeronautics and Space Administration

slide-13
SLIDE 13

Science Mission Directorate

National Aeronautics and Space Administration

transitioning unique NASA data and research technologies to the NWS

UAH UAH UAH

MODIS CTP vs. AIRS Usable Channels (2006 Dec. 4)

MODIS CTP Channels for Assimilation (%)

  • Visible agreement is seen between the MODIS CTPs and AIRS usable channels
  • Higher % of usable channels in clear regions, lower % as clouds get higher

Earth Science Office

National Aeronautics and Space Administration

slide-14
SLIDE 14

Science Mission Directorate

National Aeronautics and Space Administration

transitioning unique NASA data and research technologies to the NWS

UAH UAH UAH

CloudSat Reflectivity CloudSat Reflectivity MODIS CTP

AIRS/MODIS/CloudSat Intercomparison

AIRS psp MODIS CTP CloudSat Reflectivity Separation Point Pressure

  • Pressure corresponding to brightness temperature of first

uncontaminated channel

  • Not a CTP!

Technique sees broken clouds that neither CloudSat nor MODIS pick up

Earth Science Office

National Aeronautics and Space Administration

slide-15
SLIDE 15

Science Mission Directorate

National Aeronautics and Space Administration

transitioning unique NASA data and research technologies to the NWS

UAH UAH UAH

Conclusions

  • SPoRT AIRS assimilation focuses on short-term regional forecasts
  • Profile Assimilation
  • Prudent assimilation of AIRS thermodynamic profiles and quality indicators can

improve initial conditions for regional forecast models

  • Improvement in both T and q in over land soundings; smaller improvements in
  • ver water soundings
  • V5 profiles will be used for real time activities once on-line to generate long-

term statistics of sensible parameters and find new case studies

  • Radiance Assimilation
  • CO2 Sorting technique can be used to detect clouds and determine

uncontaminated channels in hyperspectral data with a substantial increase in usable channels over masking approach

Earth Science Office

National Aeronautics and Space Administration