Presentation on theme: "Max-Planck-Institute for Meteorology, Hamburg, Ger AeroCom … in the context of GEMS S. Kinne."— Presentation transcript:
Max-Planck-Institute for Meteorology, Hamburg, Ger AeroCom … in the context of GEMS S. Kinne
2 Overview what is AeroCom ? Goals what does AeroCom do ? Activites how does GEMS benefit from AeroCom? Initialization Evaluation
3 what is AeroCom ? AeroCom Comparisons of Aerosol simulations to DATA co-organized by LSCE and MPI-Met not officially funded (major problem) supported by global aerosol modeling worldwide AeroCom Goals document module differences investigate sub-step and sub-processes assemble useful (quality!) data-sets intensify links between groups (model, data)
4 AeroCom - Activities (1) organize regular workshops Paris 6/03, Ispra 3/04, N.York 12/04, Oslo 6/05 maintain a website Information conference summaries / papers Protocol data-format / data-request (Experiments) / input Interactive diagnosis tool Evaluations (Model vs Data) Diversity / Outliers (Model vs Model)
5 AEROCOM/DATA/surfobs.html aot SO 4 time-series local network comparisons distribution-plots scatter- plots selection menu
6 AeroCom - Activities (2) define common Experiments A: best as you can – simulation B: year 2000 with prescribed 2000 emissions* C: year 2000 with prescribed 1750 emissions* B minus C: address anthropogenic forcing INDI: sensitivity studies for indirect effects * ftp://ftp.ei.jrc.it/pub/Aerocom/ prepare useful data-sets (for data-base) Evaluate – beyond downloading (satellite combo) Combine/ Process – for added value (AERONET)
7 aot – sat. retrievals vs. AERONET …but can local data expanded in regions as here ? sat - Anet R = Anet
8 AeroCom - Questions MODELING are component modules consistent ? where is model diversity largest ? what do prescribed scenarios reveal? DATA are there data to determine skill ? are (operational, global) data available ? are data (sufficiently) accurate? can data correlations provide clues? are data applicable to scales in modeling?
9 AeroCom - Participants LO LOA 3.8/2.5yr 2000Reddy / Boucher LS LSCE 3.8/2.5yr 2000Schulz / Balkanski UL ULAQ 22.5/10yr 2000Pitari / Montenaro SP SPRINTARS 1.1/1.1yr 2000Takemura CT CANADA2.8/2.8yr 2000Gong MI MIRAGE2.5/2.0 1yr avgGhan / Easter EH ECHAM5 HAM1.8/1.83yr avg Stier / Feichter NF NCAR MATCH1.9/1.9yr 2000Fillmore / Collins OC OSLO-CTM2.8/2.8yr 1996Myhre / Isaksen OG OSLO-GCM2.8/2.83yr avgIversen et al. IM IMPACT2.5/2.0yr 2000Liu / Penner GM GFDL MOZART 2.5/2.0yr 2000Ginoux / Horowitz GO GOCART2.0/2.5 yr 2000Chin / Diehl GI GISS 4.0/5.0yr 2000Koch / Bauer TM TM5 4.0/6.0yr 2000Krol / Dentener EM ECHAM4 MADE3.8/3.810yr avgLauer / Hendricks GR GRANTOUR 5.0/5.01yr avgHerzog / Penner NM NCAR MOZART1.9/1.91yr avgTie / Brasseur NC NCAR CAM2.8/2.81yr avgMahonwald EL ECHAM43.8/3.83yr avgLohmann / Feichter all models separate by aerosol species (SU,BC,OC,DU,SS)
10 first results – model diversity differences in mass-fields are dominated by differences in aerosol processing year-to year variations are much smaller impact of streamlined emissions is minor differences among individual components (SU,BC,OC,DU,SS) are larger than for their sum data constraint usually only for (sum-) totals comp-mix diversity means absorption diversity large differences in aerosol water module (humidification) or GCM (envir) related?
11 model diversity of emission and mass Exp B Exp A Exp B Exp A emission mass emission mass
12 diversity – in aot simulations total aot diversity < aot sub-component diversity ! total aot SU aot BC aot DU aot SS aot OC aot
13 first results - data BAD data: an assimilators / evaluators nightmare do not trust given error estimates compare with quality references data of global data-sets are not globally of equal accuracy focus on regional strength, establish composites a local samples can differ from its regional value correlation can provide clues with the immediate need for absolute accuracy aerosol and other atmospheric properties
14 a case for S* (the retrieval composite) composite still no global cover!
15 AeroCom – and GEMS INITIALIZATION provide datasets on aerosol data from ground-based networks AERONET, EARLINET, EMEP, IMPROVE data derived from space sensors satellite data and retrieval composites provide reference from modeling global and complete data-sets from the AeroCom model median collaborate on aerosol emissions
16 climatology - aot / 0 / Angstrom
17 AeroCom – and GEMS EVALUATION build on AeroCom evaluation web tools diganostics and scores (e.g Taylor plots) provide a reference from global modeling statistics on simulated fields (average, diversity) provide (independent) data for evaluation quality data not used in assimilations
18 AeroCom and GEMS both activities are complementary ! AeroCom dignostic tools will provide immediate feedback on simulation performance (score ?) GEMS can build on AeroCom efforts to establish global quality data on aerosol GEMS can build on the AeroCom effort to harmonize and update aerosol emissions GEMS is expected to accelerate access to new quality data-sets for AeroCom model evalutions
19 AeroCom - current data base Remote sensing – space satellites (Modis, Misr, Toms, Avhrr, Polder …) aot (individual + composite best), Angstrom aot associated atmos properties (clouds) Remote sensing – ground AERONET (sun/sky-photometers) aot, size-dist., (ssa), Angstrom EARLINET (lidar) vertical profile, extinction In-situ ground data IMPROVE SU, OC, BC, extinction EMEP SU, PM (?) data priority for year 2000
20 General Questions to GEMS what are the priorities / GEMS needs ? …relating to AeroCom activities I can think of data aquization data assessment data integration comparisons to other modeling efforts evaluating and scoring who is going to do it / what ? extra techn. (wo)man-power for AeroCom? many GEMS participants are ready to contribute with little pieces of the puzzles. Who integrates?
22 up-scaling - of local aot 50% larger than the regional value level 50% smaller than the regional value level at GSFC: local aots are ca. 20% above regional mean with satellite data !
23 data correlations aerosol - cloud do higher cloud tops offset solar albedo losses? what process: aerosol cloud? or cloud aerosol? key a – aot A – aot (<1 m) t – cld top T L – lwc (T> 260K) x(y) – x function of y
24 model- diversity June 2004 June 2002 despite better agreement for annual global aot … large diversity in modeling remains a- aot -S sulfate ab absorption aot m- dry mass -O part.o matter w0 ss-albedo r- mee (=a/m) -B black carbon cr bc/oc ratio -N seasalt -f frac of sizes <1 m -D dust An Angstrom param. max / min factors of central 66% of aer.modules
25 clima- tology model satellite aot ssaaot with medians