OPERATIONAL AEROSOL AND DUST STORM FORECASTING AND RESEARCH

Document Sample
OPERATIONAL AEROSOL AND DUST STORM FORECASTING AND RESEARCH Powered By Docstoc
					 OPERATIONAL AEROSOL AND DUST
STORM FORECASTING AND RESEARCH


       Aerosol and Radiation Section
        Marine Meteorology Division
        Naval Research Laboratory
         westphal@nrlmry.navy.mil
      (831) 656-4743, DSN 878-4743

            Group Members:

Douglas L. Westphal     Jeffrey S. Reid
Anthony Bucholtz        Cynthia A. Curtis
Ming Liu                Elizabeth A. Reid
Annette L. Walker       Edward J. Hyer
   (ASEE)
 OPERATIONAL AEROSOL AND DUST
STORM FORECASTING AND RESEARCH



                   Outline:

1.   Model Overview
2.   Real-time observations for forecast verification
3.   Ongoing modeling/forecasting activities
4.   User needs
5.   Identify and discuss dust research issues
6.   Summary
        Multiple Forecast System Components


System                 Function
NOGAPS                 Forecasts dynamics
NAVDAS-AOD†            Data assimilation for aerosols
FLAMBE*                Detects fires, determines smoke flux
NAAPS                  Forecasts aerosols globally
COAMPS                 Forecasts aerosols regionally
FAROP**                Calculates aerosol optical properties
TEDS‡, Metcast         Navy database and distribution
† NRL Atmospheric Variational Data Assimilation System – Aerosol
Optical Depth
*Fire Locating and Modeling of Burning Emissions
**Forecast of Atmospheric and Optical Radiative Properties
‡ Tactical Environmental Data System
                 NAAPS: Navy Aerosol Analysis and
               Prediction System, An Evolving System

                                                                                 Sea salt;
                                                                                ½ degree                 Merge
             Initial Research   24 Levels    30 Levels                     T   horiz. Res.;         T     with
             Implementation                                                     10 mb top               NOGAPS
MODEL
                        Dust     Smoke                                                                        T
                                             FLAMBE                                    T      T

                                                  FAROP                        T   T

  PROCESSORS                                                                               FNMOC
                                                                                                                 T
                                                          Aerosol Data                                  Merge ADA in
                                                          Assimilation                      T
                                                                                                        NWP NAVDAS
                                                             (ADA)


                    GWI Study        PRIDE

CUSTOMERS                 www          ACE-Asia          ADAM     NAVO, ESA
                                                                                                T             T
                                                SST Dust Mask          T
                                                                                           NPOESS
                                                                UAE2                        EDR           NPOESS

                   1998               2001                 2004                    2007                  2010
11/21/2007                                        Year
 OPERATIONAL AEROSOL AND DUST
STORM FORECASTING AND RESEARCH



                   Outline:

1.   Model Overview
2.   Real-time observations for forecast verification
3.   Ongoing modeling/forecasting activities
4.   User needs
5.   Identify and discuss dust research issues
6.   Summary
     COAMPS and NAAPS Validation: Visibility Forecast
        for Hafir Al Batin, Saudi Arabia Spring 2003
                                              3/24/03                    3/25/03                         3/26/03
                              3/24/03 0:00     12:00     3/25/03 0:00     12:00       3/26/03 0:00        12:00       3/27/03 0:00
                                 -2



•3 events
                                         O bs e rv e d V is ibility




                                 0
                                 0              H AFR S D
                                          (2 8 .3 2 N , 4 6 .1 3 E )


observed at
                                 2


                                        Observed
                  vby m
                  ii i t ( )




                                 4
                  sl k




                                 4


site
                      i




                                 6
                                 8

                                 8


                                10
                                 0 12




                                12
                                 -2



                                        COAMPS
              Visibility (km)



                                  0


•COAMPS                           2
                  v by m
                  ii i t ( )




predicts
                                 4
                   sl k




                                  4
                      i




details of
                                  6
                                 8




                                  8



events                          10
                                                                        3-d ay F o r ecast o f D u st V i si b i l i ty
                                                                           H AF R S D (28.32N , 46.13E )
                                 12




                                12
                               3/24/03 0:00   3/24/03    3/25/03 0:00     3/25/03      3/26/03 0:00       3/26/03         3/27/03 0:00
                                               12:00                       12:00                           12:00
                                 -2



                                         NAAPS
                                 0




                                  0


                                  2
                 v by m




•NAAPS
                 ii i t ( )
                  sl k




                                 4




                                  4
                     i




predicts
                                  6
                                 8




                                  8


major                           10
                                                                        3-d ay F o r ecast o f D u st V i si b i l i ty
                                                                           H AF R S D (28.32N , 46.13E )


features
                                 12




                                12
                               3/24/03 0:00   3/24/03    3/25/03 0:00    3/25/03      3/26/03 0:00       3/26/03      3/27/03 0:00
                                               12:00                      12:00                           12:00




Valid Date:    24 March                                    25 March                         26 March                        27 March
Forecast Time: 00 Hour                                     24 Hour                          48 Hour                         72 Hour
                  NAAPS Validation: 14 days at Bahrain


                       Bahrain, April 26 – May 15 2006




                                                         Angstrom Coefficient (particle size) →
                     Extended Dusty Conditions
Optical Depth →




                                                                        pollution       dust
              Relevant Quantitative Verification


Verification must be done in terms that customer understands and
    needs. Optical depth and mass mixing ratio are not understood and
    may be irrelevant.
Liu et al. (2007) uses prediction rates taken from quantitative
    precipitation forecasting:
•    dust storm prediction rate: number of correctly predicted dust
    incidents/number observed dust incidents,
•    dust storm false alarm rate: ratio of number of falsely predicted dust
    incidents to number of observed clear-sky, incidents
3) dust storm threat score: (number of predicted dust
    incidents)/(predicted dust + missed dust + false alarm dust
    incidents)
4) total prediction rate: (number of correctly predicted dust incidents +
     correctly predicted clear-sky incidents)/(total observations).
                  Quantitative Validation of COAMPS Using
                            Surface Observations

                     Middle East, 16 March – 20 April, 2003
Forecast Rate →




                                                       Stations
                                                       downwind of
                                                       major dust
                                                       sources
 OPERATIONAL AEROSOL AND DUST
STORM FORECASTING AND RESEARCH



                   Outline:

1.   Model Overview
2.   Real-time observations for forecast verification
3.   Ongoing modeling/forecasting activities
4.   User needs
5.   Identify and discuss dust research issues
6.   Summary
                             Verify Vertical Distribution of Forecasted Dust

       CAPLIPSO gives us our first chance at long-term,
       global verification
                     • Example of a dust plume transported off of Africa
                     August 7, 2007
                     • CALIPSO shows dust with variable top and
                     concentration
                     • NAAPS aerosol analysis shows similar spatial
                     distribution
                     • NRL (Reid) is on NASA CALIPSO Science Team
       Now quantifying and regionalizing this comparison




                                                                  20
                                                                             CALIPSO Backscatter
                                                                             August 7 0300 Z



                                                                  10 15
                                                            Altitude (km)
                15




                             Free-Running NAAPS
  Altitude (km)




                             Dust Concentration
          10




                             August 7, 0000Z                5
   5




                                                                  0
0




                     North                         South                    North                  South
NAVDAS-AOD: NRL Atmospheric Variational Data
  Assimilation System – Aerosol Optical Depth

  Purpose:              Data assimilation for aerosols

  Status:               Running in research mode; operational next
                        year

  Input:                MODIS AOD (after quality control)

  Future input:         NPP, NPOESS, AVHRR, METOP, MSG,
                        MTSAT, AATSR, GOES-R

  Output:
   Aerosol analysis:     3-d distribution of four species

   Horizontal resolution: 1x1°
   Temporal resolution: 6 hourly
 OPERATIONAL AEROSOL AND DUST
STORM FORECASTING AND RESEARCH



                   Outline:

1.   Model Overview
2.   Real-time observations for forecast verification
3.   Ongoing modeling/forecasting activities
4.   User needs
5.   Identify and discuss dust research issues
6.   Summary
          Dust-Induced Cold Anomalies in SST Retrievals

Sea surface temperature (SST) is sensed using satellite infrared channels
    • Dust is optically active in the infrared: elevated plume appears cold
    • Analysis confirms cold bias pixels caused by large dust optical depth (AOD)
           Most current dust detection schemes can only reject data
           Unknown impact on tropical cyclone forecasting
 MODIS True Color, 12Z March 4, 2004
                                                                           MCSST Ocean QC, March 4, 2004




                                                                    2.0
                                        SST – Global Analysis (K)
                                                                    0.0




   Rejected pixels, spring 2007
                                                                    -2.0




                 Atlantic
                                                                           0                            4.5
                 Ocean                                                          Aerosol Optical Depth
               FAROP: Forecast of Aerosol
              Radiative and Optical Properties

• Post processor for calculating optical properties and
  estimating slant path visibility from NAAPS and NOGAPS
  data.
• Developed in conjunction with S. Gasso UMBC/NASA
  Goddard
• Derives extinction, absorption, and asymmetry
  parameter at all levels and optical depth at 19
  wavelengths and 3 bands (340 nm-10.6 μm)
• Forecast fields distributed via Navy database
• Operational at FNMOC; fields used daily by NAVO in
  SST algorithm
 OPERATIONAL AEROSOL AND DUST
STORM FORECASTING AND RESEARCH



                   Outline:

1.   Model Overview
2.   Real-time observations for forecast verification
3.   Ongoing modeling/forecasting activities
4.   User needs
5.   Identify and discuss dust research issues
6.   Summary
                         Dust Source Database (DSD)


• Conventional source
inventories no longer
relevant at mesoscales

• Detailed dust inventory
developed from satellite
data, weather reports,
etc., pragmatic approach

• Interannual time
variation allowed (based
on observations)

•Walker et al. JGR, 2007




   Improved dust erodible fractions required for improved COAMPS® dust predictive capability
                 Use of DSD in COAMPS



COAMPS Forecasted Mass Load (mg m-2) 0700Z October 10, 2001
          Downwind of Zabol Region, Afghanistan
             Use of DSD in COAMPS

Forecast Skill Scores:
• All improved with implementation of DSD
   OPERATIONAL AEROSOL AND DUST
  STORM FORECASTING AND RESEARCH
                    Summary
1. Model Overview
   • NAAPS, COAMPS, and Pre- and Post-processors
      are fully operational, complicated and require 24/7
      maintenance
2. Real-time observations for forecast verification
   • Needs to be done in customer’s language
3. Ongoing modeling/forecasting activities
   • Vertical distribution can now be verified
4. User needs
   • Customers are quite different from agency to
      agency
5. Identify and discuss dust research issues
   • Many exist. Need to be solved individually based
      on customer needs.