Minutes of the IOMASA Progress Meeting 3-δ,
6-7 December, 2004
held at DMI, Copenhagen, Denmark

Start of meeting: 6 December, 15:00
End of meeting: 7 December, 14:00

Participants:


IUP: 
Georg Heygster                    GH
Hendrik Laue                      HL
Christian Melsheimer              CM


DTU-DCRS: 			    
Leif Toudal                       LT
Roberto Saldo                     RS

DMI: 				    
Søren Andersen                    SA
Rasmus Tonboe                     RT 
Xiaohua Yang                      XHY
Bjarne Amstrup                    BA
Jakob Grove-Rasmussen             JG

met.no: 			    
Frank Thomas Tveter               FT
				    
SMHI: 				    
Per Dahlgren                      PD

1. Introductory items:

Welcome address.
Overview of project schedule: The 3rd project year has started, so we should start preparing publications.

2. Progress of Phase 2: Status and Results of Phase 2 of each Partner:

2.1 Part 1 (IUP): WP 2.1: Atmospheric remote sensing algorithms

(C. Melsheimer):

Total water vapour (TWV) from AMSU-B:

J. Miao' algorithm (using ratios of brightness temperature differences of 3 frequencies).

  Freq. [GHz]:  89 150  183.31+/-7  183.31+/-3  183.31+/-1
  channel no.:  1   2       3          4            5
  Channel 1, 2: Window channels, 
  Channel 3,4,5: around water vapour line.

TWV = ( C₀ + C₁ * log [(T_i-T_j-F_ij)/(T_j-T_k- F_jk)] ) * cos(theta)
4 "calibration parameters" determined from regression of radiosonde data and simulated Tbs: C₀, C₁, F_ij, F_jk
Status: relaxing saturation condition to T3-T4 > F_3,4 works very well. Can now produce
- Swath data (one TWV value for each AMSU-B pixel)
- Daily/monthly average maps, in various formats (gridded NetCDF, PostScript, PNG etc)
Swath data for years 2001/2002 almost complete. Access at http://www.seaice.dk/iomasa/vip/water-vapour, user name is partner, password: ask Roberto Saldo or Christian Melsheimer
Comparison with NCEP reanalysis data shows agreement to within about 1 kg/m², but much higher discrepancies (AMSU TWV about 3 kg/m² higher than NCEP's) at some particular places, e.g. Baffin Bay

#   XHY: Which satellites do you use?
#   CM: NOAA-15 and -16.
#   LT: There were some interferences with some search and rescue
#        transmitter in some AMSU-B channel of NOAA-15
#   CM: We should compare NOAA-15 and NOAA-16 results.
#   XHY: How is the check of remote sensing data done?
#   CM: Comparison with radiosonde data, ongoing; comparison with
#       reanalysis data fro NCEP.
#   XHY: You should rather use ECMWF reanalysis data. Check for
#        discrepancies in day/night, NOAA-15/16, cloudiness.

Surface emissivity at AMSU-A frequencies

Purpose:
- Improve assimilation of AMSU-A radiances into NWP models (IOMASA)
- Improve temperature profile retrieval near surface
Method: Use
- AMSU-A data,
- colocated radiosonde profiles
- simulated brightness temperatures with atmospheric profiles (temperature, humidity) as input
- ice concentrations from SSM/I
Then:

epsilon = (T_b(AMSU-A channel) - T_b(epsilon=1) )/ (T_b(epsilon=1) - T_b(epsilon=0)
- T_b(epsilon=1) and T_b(epsilon=0) simulated using MWMOD (MicroWaveMODel)
- atmospheric profiles (input for MWMOD) from Polarstern cruises or ECMWF model profiles
- Polarstern data only for summer; in order to have complete year-round time series, ECMWF model profiles have to be taken.
Results:
- Emissivity of sea ice (Arctic, Antarctic) and land ice (Antarctica) dependence on frequency and incidence angle: Cross-scan asymmetry visible, but to various extent.
- Time series of emissivity (monthly means): shows cycle of open water (epsilon low), new ice (epsilon very high), ice (epsilon high). Emissivity changes over Antarctic land ice as well (why?)

2.2 Part 2 (met.no/SMHI): WP 2.2 Improve Arctic high-resolution NWP

temperature data assimilation (AMSU-A)

H. Schyberg (met.no)had to cancel participation in this meeting on short notice

(P. Dahlgren):

Humidity Assimilation (AMSU-B):

AMSU-B over sea

done:
- large data set collected by EARS (Eumetsat ATOVS Retransmission System)
- quality control from AAPP (ATOVS and AVHRR Processing Package) for screening out cirrus and rain
- RTTOV7 for calculating differences between observations and NWP fields
- Reduces bias by linear regression with predictors from NWP
ongoing:
- Coding to get AMSU-B into HIRVDA (= HIRLAM variational data assimilation)
- tuning the error covariance matrix
planned:
- Test runs with HIRVDA
- Problem: under very dry conditions, surface skin temperature needed (or include skin temperature into 1D-Var assimilation beforehand)

AMSU-B over ice

AAPP algorithms for quality control cannot be used
=> use OSI SAF cloud mask
met.no code for collocating NWP fields, AMSU data, surface emissivities estimated from OSI-SAF data
=> calculate innovation vectors, study biases and how to correct them

Improved modeling of surface heat flux

(work by V. Perov)

HIRLAM set up at ECMWF
New formulation of surface fluxes implemented
3 test runs planned with 1 month of data

# SA/GH: Integrate improved surface flux modeling and ice
#        concentration retrieval

2.3 Part 3 (DTU): WP 3.2: Construction of sea ice emissivity forward model:

(L.Toudal):

Advanced statistical retrieval, R-factor and TWV algorithm

Advanced statistical (OEM, optimal estimation method) retrieval (from AMSR): SST, WS, WV, CWV, C, F, T_ice
Both TWV retrieval from AMSU-B and R-factor are also implemented; data accessible at IOMASA data browser interface
Comparison OEM TWV with AMSU-B TWV: they complement each other: AMSU-B TWV for low TWV values, OEM TWV for higher values
For OEM-retrieved quantities, ice edge/marginal ice zone causes inaccuracies most of the time because of the changing surface emissivity; AMSU-B TWV not affected by that.
Comparison OEM IC (ice concentration) with NT (NASA Team algorithm) IC:
- at low IC: OEM IC better than NT IC
- at 100% IC with FYI (first year ice) and MYI (multi-year ice): discrepancies (15%) for FYI (which algorithm is wrong?), smaller discrepancies for MYI; tie-points problem?

Time series analysis (AMSR-E, AMSU)

One year time series AMSR Tb (36,23,16,9,6 GHz) of MY ice of the GreenICE camp area:
- winter: Tb(36) < Tb(23) ... < Tb(6)
- Tb(horiz.) < Tb(vert.) at all channels, difference about 10-20 K
- June-August: large variability, because of melt ponds etc.
- In May (during GreenICE campaign): sharp drop of Tb at H-pol., all frequencies, and at V-pol., 6 and 9 GHz; probably caused by leads that open up (emissivity difference of water and ice largest at 6 GHz)
One year time series AMSU Tb (31, 89, 150 GHz) of the same area:
- as well: quite erratic behaviour June-August

#    CM: Is this really caused by surface (emissivity) effects?
#    LT: Mainly.
#    HL: But at 150 GHz, variations in summer not larger than in winter.
#    LT: They would probably be larger in summer, but the signal is
#        dampened by the higher water vapour load in summer.
#    SA: And: emissivity difference of water and
#        ice smallest at 150 GHz.
#    CM: Size of test area (GreenICE)?
#    LT: About 50 by 100 km.

In order to see why surface properties influence AMSR Tbs so drastically: Assume 100% IC, but varying snow density (100-410 kg/m³) and snow thickness (14-32 cm):
- T_ice = 256-273 K!
- OEM IC = 89-102%
=> snow layer important error source for IC retrieval

#    GH: Were snow density and thickness correlated in your simulations?
#    LT: No, just used all possible combinations of density and
#        thickness from the chosen intervals; s.th. like a worst case testing.

On ice in summer, emissivity is always a problem; thus assimilation of Tb into NWP will always run into that problem
AMSU-B TWV data might therefore be the only way to include some humidity information!

2.4 Part 4 (DMI): WP 4.2: Construction of algorithm for sea ice concentration retrieval:

(R. Tonboe):

Emission/backscatter modeling

MEMLS (snow emissivity model by C. Mätzler)
-> added sea ice (MEMLSI)
Simulated correlation between window and sounding frequencie: 23 and 37 GHz, 37 and 50 GHz, 89 and 157 GHz, 157 and 183 GHz, for various ice/snow parameters:
- Tb and polarisation difference depend on depth (position below surface) of a dense layer (crust) and a scattering layer (coarse-grained snow))
- Polarisation difference only if the crust is at the surface
Coupled thermodynamic and backscatter model: Initiated by snow and ice profile (25 snow on 5 hoar on 350 cm sea ice; site: sea ice camp in Lincoln sea, May 2004), forced by meteorological data: Comparison of modelled backscatter cross section (emissivity will follow) of snow on hoar on sea ice with Seawinds scatterometer data: O.K.

(S. Andersen):

AMSR-E

AMSR-E: Comparison of data from 2 sources
1. BUFR data (level 1A) from NESDIS (2-3h delay)
2. HDF data (level 1B)
Offset (difference) 0.5 K
Scan A / Scan B difference at 89GHz: calibration mismatch of A and B horn in level 1 and 2; polarisation difference not affected very much for typical sea ice case

#  GH: Contact Frank Wentz.
#  LT: Additionally: oscillation of Tb even within just one scan, A or
#      B (reason: total power radiometer, calibrated once per each
#      scan, but still some oscillation)
#  GH: Could we also get the data from NESDIS?
#  SA: We'll have to check.
#  GH/SA: There are a number of questions to JAXA and RSS (F. Wentz)

SAR validation data

15 scenes processed, ESA data finally coming in!
Accuracy: Comparing result from two independent operators: Ice/water confusion 2.1%

Ice concentration system

Combine virtues of 89 GHz and low resolution algorithms
Atmospheric data from other space platforms should be integrated via assimilation inNWP model
Flag to describe surface and atmospheric conditions (R-factor, emissivity anomaly from thermodynamic model)
No general merging method

To Do (final stage of project)

Analysis of SAR IC versus radiometer IC.
Selection of high and low frequency algorithms
IC system with atmospheric correction
Example: AMSU-B TWV versus HIRLAM TWV (1 Jan, 2001): one low pressure system (moist air) over Arctic ocean absent from HIRLAM output.

#  XHY:  This could also be a problem of the satellite data.
# SA/CM: Maybe, but the time evolution of that system over several days
#        looks very reasonable, as looks the time evolution of other systems
#        over the Arctic (moving in from Atlantic, eastward
#        displacement etc.)
# SA/GH: We should search systematically for discrepancies between TWV
#        from AMSU-B and from HIRLAM (possible weather systems missed by
#        HIRLAM)

3. Review of Phase 2 / Inter-task Communication

Input needed by met.no:
- Since coding is done, emissivity is needed for test runs (using SAF ice concentration; emissivity forward model is then needed)
- DTU can provide that
Input needed by SMHI
- AMSU-B emissivities (coordinate with IUP)
- Some info on emissivities and their correlations (from IUP's participation in SEPOR/PolEx, N.Selbach's work) will be made accessible to IOMASA partners (member website).
Input from SMHI/met.no needed by the other IOMASA partners:
- HIRLAM data, 2001/2002 (Integrated quantities or profiles, depending on purpose)
- Should be ready/accessible (see PM2 or MTR minutes, ask Harald)
DMI to coordinate/interact with SMHI in order to include IOMASA IC in modeling (heat flux)

#  LT: Algorithms: prefer C or so to IDL.
#  CM: You could us your C implementation of our (IDL) algorithm for
#      checking...

"Real time demonstration" towards the end of the project: We must start getting ready for that. Some brainstorming results:
- Action Item for PM3 in March 2005: We must decide on details of demonstration phase: first see what is possible.
- "(near) real time" here means the time lag can be about 3 hours.
- As to HIRLAM data: Now ready: 2 years (2001/2002); for real time demonstration we will hopefully have data from the HIRLAM version that assimilates AMSU data.
- There might be constraints because of the very large data volume.
- Who - except the EC officer - do we want to invite to the demonstration?
- Before deciding that, we have to know what exactly we are going to demonstrate.

Liquid water retrieval over ice with SSMIS?

H. Laue works as a visiting scientist (half a year) for DMI.

Probably no SSMIS data within this project
What to do?
1. model simulations for all SSMIS channels
2. use data from SSM/I, SSM/T2, AMSU, AMSR ??/li>

#    LT: You use radiosonde profiles for modelling; CLW from where?
#    GH: profile (temp.erature, humidity) + cloud model = CLW (Liza
#        Zabolotskikh's work at IUP)
#    LT: You said you have only 37 suitable radiosonde profiles (out of
#        thousands). Why?
#    HL: Requirement: cloud cover and 100% IC
#    LT/SA: 100% IC not necessarily required.
#    HL: What is the purpose of the 60.3 GHz channels?
#    CM: Ask ARTS spectroscopy people.
#    SA: Do you plan a sensitivity study with emissivity?
#    HL: Yes.
#    CM: Emissivity correlations: AMSU-B: Nathalie Selfish; AMSU-A: Nizy
#        Mathew.

Project management

Reports, Publications

For a publication to counted as output of IOMASA, the IOMASA contract number must be mentioned in the acknowledgments
Annual Report: As last year, see also e-mail from a few weeks ago.
Deliverable 3.2.1, 3.2.2 ("report and code"): LT and RT
Annotated list of due deliverables will be e-mailed to all by Christian
TIP (Technology Implementation Plan): login and passwords of each partner will be communicated by Christian

Planning of PM3, PM4

half day: Just the partners

another day: with UAG

#  CM: Please look at the UAG comments from PM1 (on IOMASA member
#      website)

Presentations to UAG should focus on:
- What has been accomplished
- Internal interaction
- Interaction with outside
Time and place: 14-15 March, 2005, DTU, Copenhagen. Starting Noon 14 Mar (Monday), UAG to join morning 15 March (Tuesday), till afternoon or early evening.
PM4: Probably in Oslo, week

6. The Future

6.1 DAMOCLES

DAMOCLES = Developing Arctic Modelling and Observing Capabilities for Long-term Environmental Studies
FP6 (Sixth Framework Programm) project
Pre-Proposal submitted in October 2004
If accepted, full proposal is due end of Feb 2005.
includes atmosphere part, here IOMASA is relevant
4 years project (incl. field work)
about 50 partners
entire budget about 15 Mio. EURO, thus roughly 200,000 EUR per partner, i.e. about 1 (1/2??)PostDoc

6.2 IPY (international polar year)

not for funding, but being an IPY project will greatly help getting national funding.
Dealdline: Jan 14, 2005; 400 word abstract/letter of intent.
We (IOMASA)should write a letter of intent.

#  GH: What to put in?
#  SA: Partners in USA/Canada (e.g., International Ice Charting Working
#      Group)
#  FT: Lars-Anders Breivik suggests to coordinate writing of the letter.
#  GH: Will contact L-A Breivik. - Ice, Ocean, Atmosphere.

G. Heygster writes a draft letter (this week), comments by IOMASA members by next week.

#  LT: Remember to check the (rather strict) form for the letter!
#      Available at international IPY web site.

DAMOCLES group will probabley also wrte a letter of intent (DAMOCLES meeting 8 Dec at Bremerhaven)

7. Any Other Business

7.1 EuroGOOS meeting, June 6-9, 2005

each partner shoud think about an individual contribution: please tell Georg by Dec 10 if you consider an individual presentation
if not: paper/presentation of the whole project (G. Heygster, input from all)

GLOSSARY/ACRONYMS:

AAPP	ATOVS and AVHRR Processing Package
AVHRR	Advanced very high resolution radiometer (Vis./NIR sensor on NOAA satellites)
ATOVS	Advanced TIROS Operational Vertical Sounder (passive microwave sensor on NOAA satellites)
CLW	cloud liquid water
CWV	column water vapour (= TWV = PWV)
DMI	Danish Meteorological Institute
DTU-DCRS	Technical Univ. of Denmark, Danish Center for Remote Sensing
FP	Framework Programme
FY	first-year (ice)
HIRVDA	HIRLAM variational data assimilation
IC	ice concentration
ICEMON	Sea ice monitoring in the polar regions
IPY	International Polar Year
IR	infra-red
IUP	Institut für Umweltphysik (Environm. Physics), Univ. Bremen
met.no	Norwegian Meteorological Institute
MIZ	marginal ice zone
MY	multi-year (ice)
NCEP	National Centers for Environmental Prediction
NT	NASA TEAM (algorithm for sea ice concentration retrieval)
NT2	NASA TEAM 2 (algorithm for sea ice concentration retrieval)
NWP	numerical weather prediction
OEM	optimal estimation method
OSI SAF	SAF on Ocean and Sea Ice
PWV	precipitable water vapour (= TWV = CWV)
RSS	Remote Sensing Systems (comapany headed by F. Wentz, www.ssmi.com
SAF	Satellite Application Facility
SMHI	Swedish Meteorological and Hydrological Institute
SST	sea surface temperature
T	temperature
Tb	brightness temperature
TCW	total cloud water
TWV	total water vapour (= CWV = PWV)
WV	water vapour

Minutes prepared by Christian Melsheimer

Minutes of the IOMASA Progress Meeting 3-δ, 6-7 December, 2004 held at DMI, Copenhagen, Denmark