game-jp の皆様 檜山@名古屋大・管理者 です
下記のメ−ルが北大の藤吉先生から投稿されましたが、うまく流れませんでした。
代理投稿いたします。
**************************************** 以下、藤吉先生のメ−ル
>
> 複数受け取られる方、お許し願います。
>
> GEWEX Cloud System Study (GCSS)では、
> 様々な観測事例についてモデルの比較実験を
> 行っています。その結果については、
> Mon.Wea.Rev. やBull.Amer.Met.に既に
> いくつか報告されています。
>
> 今回、新たな比較実験の提案が届いておりますので
> 以下に添付いたします。興味をお持ちの方は、
> 是非ご参加下さい。
>
> また、新たな巻雲観測計画(CRYSTAL)の最終版も
> できたそうですので、ご覧下さい。
>
> GCSS-WG1:
> The ARM shallow cumulus over land case is now ready for first
> release.
> To obtain details either
> ftp
> connect email.meto.gov.uk (login as "anonymous")
> cd pub/apr/arm
> or URL:
> ftp://email.meto.gov.uk/pub/apr/arm
>
> The following files will be found
> ARM_README.TEX List / description of files in directory
> CASE_31_MAR_99.PS Case description
> QSAT_CODE.FOR Fortran for saturation mixing ratio calculation
> BC_CODE.FOR Fortran for obtaining surface stress
> RESULTS_31_MAR_99.PS Preliminary results (UKMO, MPI) for time series
>
> Please let me know of any problems you encounter, either with getting
> and reading the files, or with the case itself.
> Andy
> ------------------------------------------------------------------
> Andrew R. Brown
> Met O(APR)
> Meteorological Office, Room 258
> London Road,
> Bracknell,
> Berkshire.
> UK. RG11 2SZ
>
> email: arbrown@meto.gov.uk
> fax : +44 (0)1344 854493
> phone: +44 (0)1344 856461
> ------------------------------------------------------------------
>
>
> DRAFT PROPOSAL FOR WG3 FASTEX IOP16 INTERCOMPARISON
> ===================================================
>
> 1. OVERALL AIMS AND METHODOLOGY
> The overall aim of the intercomparison will be to improve the
> parameterisation of clouds and cloud processes in extratropical
> cyclones in GCMs. The approach to doing this is to run model simulations
> of the system at a number of resolutions and compare the representation
> of the clouds and cloud related structures both between models,
> resolutions and with observations.
>
> In order to elucidate the reasons for differences in the representations
> it will be important to understand differences between the cloud and
> microphysics schemes in the various models. A fundamental question in
> the representation of cloud is the behaviour of the cloud scheme. In the
> case of diagnostic schemes, this amounts to how much cloud/cloud cover
> is produced by a given q etc., and the property to be tested and
> hopefully improved is how much each scheme is able reproduce the correct
> average cloud properties at different model resolutions. In the case of
> prognostic schemes, where various source and sink terms represent the
> rate of change of some chosen cloud properties, the aim is to
> investigate both the structure of the formulation and the accuracy of
> the model forcing terms (e.g. sub-gridscale fluxes). This implies that
> the study should concentrate as much on the parametrizations of sub-grid
> fluxes of heat, moisture etc. within low-resolution GCMs as on the
> representation of cloud per-se.
>
> This intercomparison should provide data for a direct test of various
> schemes at various resolutions. An additional tool in a second
> stage of the intercomparison will be one or more sensitivity tests
> to various model cloud scheme/microphysical parameters, for example
> response of cloud distribution to q etc., ice fall speed, vertical
> resolution etc..
>
> 2. INTRODUCTION TO IOP16
> FASTEX IOP16 was a case of a rapidly moving, rapidly developing
> frontal wave secondary cyclone. The system developed
> from being a trough to a well developed system in the
> period 0Z to 12Z on 17th Feb 1997. The aircraft measurements
> (dropsonde, radar and aircraft level) were made in the period
> 6Z-12Z. A feature of this system is multiple cloud heads which
> emerge from under the cloud shield starting at around 6Z and
> a focus of interest may be the source of the instabilities which
> lead to the circulations which generate these.
> An important feature of this case, from the modelling point of
> view, is the rapid speed of movement of the system which is about
> 30m/s. This may impose a constraint on how long the system can
> be modelled for depending on the domain size. To some extent this
> is, however, ameliorated by the rapidity of development which shortens
> the time span of interest. A second feature is that the flight strategy,
> which aimed to map out different parts of the system, lead,
> fortuitously, to an approximately constant group position for the centre
> of the aircraft tracks, thereby making direct comparison with aircraft
> data simpler.
> General information about IOP16 and details of the various observations
> taken can be found at http://www.met.rdg.ac.uk/FASTEX/i16sum.html
>
> 3. LAM/CRM SPECIFICATIONS
> The input data will be from runs of the UKMO Unified Model (UM)
> at 0.105 deg (approx 12km), 45 level resolution. The grid for
> this model was (positive longitude=east of Greenwich meridian):
>
> rotated lat/long grid
> 394x247 points, 0.105x0.105 degree resolution
> pole lat/long: 34.5N,150 degrees
> top left gridpoint lat/long: 10.22,-23.24 on rotated grid
>
> This domain is the whole area illustrated in the figures, which also
> shows, in blue, the approximate C130 tracks. This model has been run
> starting at 9Z on the 16th Feb from 50km operational model data. It
> was run forward from this time using operational 50km analysis
> boundaries and including 4 3 hourly assimilation cycles (12,15,18,21Z
> on the 16th).
> Some of the basic fields from this run are posted on the web for
> reference at http://www.met.rdg.ac.uk/~sws97hwl/GCSS/index.html
> Model dumps will be provided at 0Z 17th and onwards at 3hrly intervals
> in order to provide starting and boundary data. The data will be
> provided on a 0.21 degree (approx 24 km) grid, which may be reduced by
> one grid box to allow for grid staggering.
> The GCSS simulations will include nested simulations at four different
> (approximate) resolutions: 24km, 12km, 4km and 2km.
> The 24km simulation should cover the time period 0-12Z.
> The suggested 24km grid is shown in purple on the figures and is:
>
> rotated lat/long grid
> 155x109 points, 0.21x0.21 degree resolution
> pole lat/long: 34.5N,150 degrees
> top left gridpoint lat/long: 8.77,-16.0 on rotated grid.
>
> If it is desired to run on a non-rotated or other type of grid a
> similar grid may be defined subject to the system track from
> 0-12Z being covered, and a suitable surrounding area (in particular
> the cloud head at 12Z should be included) and, of course, the
> whole grid being inside the area of the supplied data. The model
> data posted on the web should help the defining of such grids.
>
> The suggested 12km grid, for those running on a rotated grid, is:
> rotated lat/long grid
> 150x150 points, 0.105x0.105 degree resolution
> pole lat/long: 34.5N,150 degrees
> top left gridpoint lat/long: 7.0,-2.0 on rotated grid.
> This is shown in red on the figures.
> Once again alternative grids may be used. In this regard it should
> be noted that the FASTEX-CSS simulations are already underway, using
> ARPEGE re-analysis data at 0.4 degree resolution to drive a simulation
> on the following grid:
> NON-rotated lat/long grid
> 150x150 points, 0.1 (latitude) x 0.17 (longitude) degree resolution
> top left gridpoint lat/long: 60.55,-25.835
>
> This is shown on the figures as the outermost boundary of the green
> area.
> The simulation data will also be averaged to 1 (latitude) x 1.7
> (longitude) degree (approx 100 km ) resolution. these 1 degree boxes
> are shown. Unfortunately it will not be possible to run with this
> grid since current data does not cover it all (as can be seen from
> the figures) but a similar one could be used.
>
> The 4km and 2km CRM runs should be centred on the position of the
> aircraft tracks, 55N,16.4W with the axes aligned to LOCAL E-W, N-S.
> The range of these runs should be at least 6-12Z (possibly starting
> earlier). The 2km run will provide an approx 300x300km domain which
> will correspond to roughly the central light blue box in the figures.
> It should be noted that this area is very small when compared to the
> speed of motion of the system (it takes the system only about 2 hours
> to travel from one corner to the opposite one). In the light of
> this is it worth carrying out 2km simulations at all unless larger
> domains can be used??
>
>
> 4. GCM SPECIFICATIONS (needs to be added)
> It should be noted that there may be problems in obtaining a good
> representation of the system with global climate resolution models.
> I attempted to model the system starting from both ECMWF and
> UKMO global data and found that it did not develop satisfactorily.
> What do people think about the need for this?
>
> 5. SCM SPECIFICATIONS (needs to be added)
>
> Suggest SCM simulations of innermost, 300 km domain, driven by
> tendencies from GCM at grid centre. We can provide tendencies, or
> modellers calculate their own. What do people need?
>
> 6. OUTPUT FIELDS
>
> (a) Box average intercomparisons.
> It is proposed that, whatever resolutions are run, the data analysis
> concentrates on the approx 300x300 km box almost centred on the C-130
> track: this should (approximately) be the domain of the 2 km CRM
> runs, centred at 55N,16.4W, with the axes aligned to LOCAL E-W, N_S.
> All models should provide average data for this gridbox (covering
> the area 53.5N-56.5N,18.95W-13.85W). The larger scale models should
> also provide data for the 8 surrounding boxes (shown light blue),
> interpolated (if necessary) to a true 3(latitude) x 5.1 (longitude)
> grid box. This will enable 'large scale' gradients between boxes
> to be assessed, along with variability. (For the 24 and 12 km models,
> it would obviously be preferable to interpolate first to a fine,
> 0.1x0.17 degree non-rotated lat/long grid, then average up).In
> averaging from high to low resolution, relative gridbox area should
> be taken into account (by weighting by cos(latitude)). The exact
> specifications of the boxes are that they are bounded by the following
> lattitudes and longitudes:
>
> 59.5N,56.5N,53.5N,50.5N
> 24.05W,18.95W,13.85W,8.75
>
> Thus, the top left box, box 1, covers 56.5N-59.5N, 24.05W-18.95W
> the bottom right box, box 9, covers 50.5N-53.5N, 13.85W-8.75W
>
> Box numbering is :
> 1 2 3
> 4 5 6
> 7 8 9
> where any numbering is required.
>
> Include basic eddy fluxes of heat, moisture, momentum, Q1, Q2, w/omega,
> q,t,u,v,pmsl, precipitation rate + microphysical quantities. Larger
> scale (i.e. non-CRM) models should include parametrized fluxes
> (primarily convection and BL schemes) as separate outputs.
> Since the box averaged data will be relatively small it should be
> acceptable to exchange it as ASCII. The ordering (boxes, levels,
> variables, times) must be made clear. The pressure or geopotential
> heights of the levels used will need to be provided.
>
> For definitions see http://www.tor.ec.gc.ca/GEWEX/GCSS/largesc.htm.
> Add any parameters required by (or predicted by) cloud schemes.
>
> (b) Gridded Data
> It should be noted that we expect that, at least initially,
> most of the quantitative work on the data will be carried out on
> the box averaged data discussed in the last section. The gridded
> data will be required in order to compare the evolution of fields
> between different models. In order to facilitate quick comparisons
> (eg overlaying fields) we would like to specify a 90x90 grid
> which corresponds to the area of the 9 blue squares in the figures
> for participants to interpolate their gridded data onto. The
> specifications of this grid are:
> non-rotated lat long grid
> resolution 0.17 (E-W)x 0.1(N-S) degree
> Top left point is 59.5-0.1/2=59.45N, 24.05+0.17/2=24.135W
> Data output from top left (NW) point first, scanning across W-E then
> N-S.
> Data output as 32 bit IEEE reals, each 2D field preceded by a 64 word
> header. Format and code for writing to be supplied by UKMO.
>
> The details of the variables required are:
> 3hrly output
> Single level: pmsl, T*, surface sensible/latent heat fluxes.
> Multiple level (25mb resolution): z,T,u,v,q - also mircophysical
> quantities: qcl, qcf, rain(+ snow if separate variable), graupel,
> condensation/evap of water/ice/rain/snow, autoconversion of
> qcl into rain, autoconversion of qcf into snow, collection of
> qcl by rain, collection of qcf, qcl by snow, Bergeron-Findeisen
> mechanism, melting/freezing of cloud condensate, melting/freezing
> of precipitating particles.
>
> Validating observations will be C-130/P3 aircraft data (including
> dropsondes, radar, flight level microphysics), possibly upsondes
> from coastal stations.
>
> In addition to the above gridded data it would be helpful if
> participants could provide us with one or two basic fields (pmsl
> others??) 3hrly over the full area of their domains in the form of
> image files (gif etc) to give an initial approximate idea of the
> evolution in each model.
>
> (c) ISCCP
> cloud fraction (plus separate ice and water fractions where possible)
> cloud top pressure
> column cloud optical depth
>
> 7. MICROPHYSICS TESTS
> To be carried out after the initial runs AFTER we have ascertained
> exactly what microphysics is in each model and with that knowledge
> come to a decision as to what is sensible.
> Suggestions so far include sensitivity to ice fall speed, lambda
> and vertical resolution. Obviously parameter space is very large
> so we need a definite strategy for this before embarking.
>
>
> FIGURES
> Show IR images as background over area of UKMO 12km simulation
> at 6Z and 12Z on 17th Feb 1997. Grids marked as described in text.
> These are attached as gzipped postscript. (If you have any trouble
> with the attachment they are also on the web at the URL mentioned
> in section 3)
>
> --
> Humphrey W Lean,
> Mesoscale Modelling Group,
> Room 2L62, Joint Centre for Mesoscale Meteorology,
> Department of Meteorology, University of Reading,
> PO Box 243, Reading RG6 6BB, UK.
>
> Tel: 0118 931 6624 (University ext 6624, GTN 1441 6624)
> Fax: 0118 931 8791 email: hwlean@meto.gov.uk
>
>
> <file://d:\progra~1\eudora-j\attach\FASTEX_GCSS06.ps.gz>1bb9a792.jpg<file:
> //d:\progra~1\eudora-j\attach\FASTEX_GCSS06.ps.gz> FASTEX_GCSS06.ps.gz
>
>
> <file://d:\progra~1\eudora-j\attach\FASTEX_GCSS12.ps.gz>1bb9a7f0.jpg<file:
> //d:\progra~1\eudora-j\attach\FASTEX_GCSS12.ps.gz> FASTEX_GCSS12.ps.gz
>
>
> *********************************
> The final version of the CRYSTAL Research Plan has been released on the
> FIRE website (http://asd-www.larc.nasa.gov/fire/index.html).
>
> NASA will soon issue an NRA (solicitation for proposals) to form a CRYSTAL
> Science Team.
>
> =========================================================================
> = David O'C. Starr email: starr@climate.gsfc.nasa.gov =
> = Code 913 express delivery: Bldg 33, Rm C-308 =
> = NASA Goddard Space Flight Center voice: 301-614-6191 =
> = Greenbelt, MD 20771 fax: 301-614-6307 =
> =========
***************************************
*藤吉康志 *
*寒冷海洋圏科学部門 *
*北海道大学低温科学研究所 *
*〒060−0819 札幌市北区北19西8 *
*Tel. & Fax(1) 011−706−5491 *
* Fax(2) 011−706−7142 *
*e-mail fujiyo@lowtem.hokudai.ac.jp *
*http://stellar.lowtem.hokudai.ac.jp/ *
***************************************
************************************************************************
檜山哲哉 〒464-8601 名古屋市千種区不老町 名古屋大学大気水圏科学研究所
TEL:052-789-3478 FAX:052-789-3436
e-mail:hiyama@ihas.nagoya-u.ac.jp
************************************************************************