Summary of the POET September meeting

September 15, 2000

Cambridge

 Attended the meeting:

 Ivar Isaksen, University of Oslo (ivar.isaksen@geofysikk.uio.no)

Stig B., University of Oslo (stigbd@geofysikk.uio.no)

Bojan Bojkov, NILU (bojan.bojkov@nilu.no)

John Burrows, University of Bremen (john.burrows@gome5.physik.uni-bremen.de)

Nick Savage, University of Cambridge (nick.savage@atm.ch.cam.ac.uk)

Kathy Law, University of Cambridge (kathy@atm.ch.cam.ac.uk)

Valerie Gros (vgros@mpch-mainz.mpg.de)

Jos Olivier, RIVM (jos.olivier@rivm.nl)

Cathy Clerbaux (ccl@aero.jussieu.fr)

Claire Granier (clg@aero.jussieu.fr)

 Administrative issues:

Claire gave a summary of the meeting which took place in Brussels in June at the European Commission, which discussed the clustering of POET with other European projects. POET is now part of a cluster with the TROTREP, FUTURE_VOC and SUB-AERO projects :

TROTREP : Tropospheric ozone and precursors, trends, budget and policy.

Coordinator : P. Monks, Univ. Leicester, UK

Participants : Univ. Berne, IVL Gothenburg, TNO Apeldoorn, Norwegian Meteo Oslo, NILU, Univ. Utrecht)

FUTURE_VOC : BVOC emissions of european forests under future CO2 levels : influence on compound composition and source strength

Coordinator : Steinbrecher, FHG Garmich, Germany

Participants : Univ. Basel, CNR Rome, Instituto Miglioramento Genetico Alberi forestali Firenze, Lancaster Univ.

SUB-AERO : Subgrid scale investigations of factors determining the occurrence of ozone and fine particles

Coordinator : Lazaridis, NILU, Norway

Participants : Univ. Athens, NCSR Demokritos Athens, Univ. Essex, Academy of Sciences Czech. Republic, JRC Ispra.

2 rapporteurs will have to write reports (when ?) for each clusters. The rapporteurs for this cluster are M. Millan (Spain) and Claire.

• NOTE : a first detailed yearly report for the POET activities is due April, 30 2001. Claire will send the informations on the report in March 2001, so that we can start to write.
  

•  Next meeting : in Bremen April,2 2001. We should complete the yearly report during the meeting. We would like also to invite the coordinators of the other projects in our cluster + the coordinator of MOZAIC (à meeting of 1.5 to 2 days ?)

Workpackage 1: Distribution of ozone precursors from satellite data

 John discussed version 1.0 of the NO2 tropopospheric column data, which are now available from January 1997 to June 2000. He showed the average NO2 tropospheric columns and 1997 and 1998, discussed some regional features corresponding to anthropogenic emissions, lightning and biomass burning emissions.

 People who would like to have access to the data should contact Andreas Richter.

 He discussed the methods used to separate stratospheric and tropospheric ozone, and how to better take into account the airmass factors, the clouds and the aerosols. He proposed new approaches, and showed how for example the use of 2 different wavelenghts could improve the overall accuracy of the results.

 He showed also results concerning formaldehyde, tropospheric ozone and SO2.

 = = = = =

 Cathy give an update of the retrieval of CO tropospheric columns from IMG. The data for the June 16-19 1997 period are ready, together with their accuracy. Data for April 1997 are also ready. The retrieval of ozone tropospheric columns for June 1997 is in progress, and the retrieval of methane should start in a few months.

• It would be great if Cathy and John could put on the Web page a figure showing NO2 and CO for the same period (june 1997).

 Cathy showed a figure giving a comparison between assimilated CO for June 1997 and the corresponding surface data from CMDL. The agreement is good, except for the high southern latitudes, where the CMDL values could be underestimated by about 15%.

Workpackage 2: Emission inventories

 Jos gave a presentation of the current status of the EDGAR inventory of anthropogenic emissions. The EDGAR v2.0 has been updated, some categories have been added, emission factors have been updated. The spatial/temporal distribution of biomass burning has been included. The inventories by country will be ready at the beginning of October.

 Jos also gave indications on the changes of anthropogenic emissions for some species from 1970 to 1995. He also gave some indications on the strong seasonality of some emissions such as emissions from the residential sector.

• Question : What do we want for the timing of the spatial distribution of biomass burning emissions ? We agreed that we should first focus on 1997, which is the year for which we have most of our satellite data.

 By mid-december 2000, the 1995 new Edgar will be available on a 1x1 degree grid.

 = = = =

 Jos and Jean-François discussed prior to the meeting about natural emissions, to try to agree on an emission database which will be used for the intercomparisons of the models. A few indications on what they agreed on :

 Methane emissions : Use Matthews and Fung

 NOx from soils: Use the GEIA inventory and scale it to a total emission of 8 Tg N/yr.

 NOx from lightning : Use Price and Penner distribution and scale it to 4 or 5 Tg N/yr (needs to be decided)

 CO from vegetation : distribution based on the npp, and scaled to a total of 160 Tg CO/yr.

 Soil sink : proposition JF :  vd=0.05*NPP_m, where NPP_m = monthly NPP

 isoprene : still need discussion, use GEIA (503 Tg C/yr ?)

 terpenes : from GEIA

 = = = =

 Work within TROTREP

 Jos, who is also part with TROTREP (see at the beginning), showed us what is planed within TROTREP.

 The history of atnthropogenic emissions from 1970 to 1990 will be determined, together with the time profiles. There will be sensitivity studies to test how models are sensitive to the seasonality of emissions, as well as to the altitude of the emissions.

Workpackage 3: Intercomparison of CTMs

 4 models are participating in POET: TOMCAT from Cambridge, OSLO-CTM2 from the University of Oslo, MOZART from CNRS in Paris, and IMAGES from BISA, Brussels.

 The models are not yet fully ready for the planned intercomparison, but a preliminary comparison of the model results has been done. Two types of runs have been done, one simulating the distribution of Radon (emissions as indicated in the Jacob et al., J.G.R. 1997 paper), and one with a stratospheric tracer, with an initial condition of 1 in the stratosphere, and 0 in the troposphere in January, and the distribution of the tracer at the end of March is archived.

 Radon simulation : The IMAGES (average meteorologys), CTM2 (1996 ECMWF) and MOZART (GCM meteorology) we compared, and we looked at the surface and 300 hPa distributions in March and July. The distributions were similar, but we need now to plot them with the same color scale for better comparisons.

 Stratospheric tracer : Only IMAGES and CTM2 runs were performed, and they show rather different results, with a larger downward transport in IMAGES than in CTM2.

 Stig presented a detailed update of the Oslo-CTM2 model, which will be used for the intercomparisons.

Discussion of the future work :

•  In the runs we will perform during the following months, we will all use the ECMWF analysis for 1997, at 31 levels.
  

• We will start to convert our ouput into the HDF format (see below) and send them to the NILU database.

Radon :

•  We need a more detailed analysis of the maps, looking at regional differences, at observations of Radon (as in the Jacob et al. Paper)
  

• In the yearly report, we will have to state more clearly the objectives of the intercomparison of these tracers : for example, it is important to test the convection schemes with the Radon tracer as convection is a very important process, more particularly over the tropics in the biomass burning areas, which represent a large source of CO.
  

•  We should also do the same runs of Radon, without convection from May to spetember.
  

•  We should archive time series, in order to be able to calculate standard deviations

Stratospheric tracer :

  We will perform 4 series of runs, each being 3-month long. Each of the runs will start on  the 1^st of January, April, July and October, and archive the last 10 days of the 3^rd month.

Workpackage 4: Inverse modeling

 Claire gave an update of the work done on inverse modeling. This work was performed using the IMAGES model, and it has focused on the inverse modeling of CO using the CMDL observations of CO over the past few years. Several asumptions are made in this first step : errors on model results and observations are known, they are gaussian and indépendant, and there exist a linear relationship between CO emission and concentration. In this study, only direct surface emissions of CO were optimized, which means it was assumed that there is no error on the production of of CO from hydrocarbons oxidation. The optimization of surface emissions was done over 12 large areas, assuming a fixed distribution of emissions inside these areas.

 The first results are very encouraging. During the next months, the results of the MOZART model will be used, satellite data will be included in the inverse modeling process.

 The development of the adjoint of MOZART will start soon. The adjoint generator TAMC will be used for these developments, in collaboration with people from NCAR (Boulder, USA) and from the Max Planck Institut of Meteorology in Hamburg).

Workpackage 5: Isotopes

 Valerie gave an overview of the origin of CO and CH₄ isotopes and of currently available measurements. There are 10 stations measuring CO isotopes, 3 have stooped making measurements a few years ago. The 7 ongoing stations cover a latitude range from 28N (or S ?) to 79N.

 Measurements are also available from a few campaigns (INDOEX), from the transiberian campaigns (TROICA : 1992, 1997 and 1999), from the 25 flights from 1997-2000 within the CARIBIC project (flights from Germany to the Maldives). Other institutes have also some surface data, like at the Barbados, and some aircraft are also available.

 Valerie gave also a list of recent model studies on CO/CH4 isotopes.

 The data will be included in the POET database soon.

 Workpackage 7: tagging of ozone

 We talked quite a bit about the ozone coloring, as we have not been able to solve the most difficult problem : How do we deal with the recycling of species.

 Coloring of ozone : rather straightforwars, as sources of ozone precursors from the stratosphere are almost negligible.

 
We discussed several possibilities :

 1.   tagging of ozone precursors : NOx, CO, CH4 and other HCs. This step is necessary, and we should start with this during the coming months. We will have to agree on what we want to tag (area, type of emission, ..)

    Nick will propose a list of source areas and processes so that we can start this  agging.

2.   Use families such as Ox and NOx in order to reduce the difficulties caused by the cycling of species. This may mean different models with different approaches.

3.   Consider the regions one at a time to reduce demands on memory.

4.   Use of budget code with tagging of precursors (Cambridge)

5.   Multiple origin problems (for example, one anthropogenic NOx reacting with a natural RO2). Possibilities :

a)   split ozone 50/50 between categories

b)   create N² number of regions (N=number of regions or type of emission)

c)    consider that, for exmaple, ozone coming from one european NO2 + one asian RO2 is the same ozone as one asian NO2 + one european RO2.

The POET database :

The location of the data base at NILU is zardoz.nilu.no (IP : 128.39.104.1)

POET directory tree :  /nadir/projects/other/poet

To access the data in the database, we need to sign the data protocol.

The data protocol will be sent to all people in the POET email list  (in the attached file).

The access will be retricted, we have to think about what we do with people in the same cluster.

Bojan gave a presentation of the new Nadir data base :

•   Strict metadata guidelines

•    Model data will have to be defined correctly

•    HDF 4.1 r3 format

•   Inventory through a relational database

HDF format : information (freeware) at :

http ://hdf.ncsa.uiuc.edu/hdf4.html

 or through nadir.nilu.no/calval

 HDF is a binary format, platform indépendant, and there is a suite of tools available.

 
Informations on metadocuments :

nadir.nilu.no/calval/documents.htm

   We have to start to transfer our archive into HDF and into the POET database.

Workshop on emissions of chemical species and aerosols :

Dates of the workshop : June 19-22  2001 in Paris

We discussed and improved the preliminary program : there is now a full day concerning available observations and their use for a better quantification of emissions. A first circular has been sent to the non-POET people to have an idea about the number of people who might be interested, and a call for abstracts will be released in December, for a submission deadline of March, 31^st.