Agenda Supporting Items

4km GOES 8/9 data to universities

Unidata Support

 Date: Tue, 22 Apr 1997 13:33:14 -0600 From: Unidata 
      Support  Reply-To: usercomm@unidata.ucar.edu, 
      support@unidata.ucar.edu To: gembud@unidata.ucar.edu, ldm-users@unidata.ucar.edu, 
      mcidas-os2@unidata.ucar.edu, mcidas-x@unidata.ucar.edu Cc: support-announce@unidata.ucar.edu 
      Subject: 970422: LDM/IDD implications of 8-to-4 km resolution changes for 
      GOES imagery *** Be sure to read the section concerning LDM product queue 
      size below *** Users of the Unidata-Wisconsin Datastream: At 0 Z Wednesday, 
      April 23, (6 pm MDT tonight!) SSEC and Unidata will be increasing the spatial 
      resolution of the GOES-8 Infrared image product (McIDAS product code UI) 
      from 8 km to 4 km. The GOES-8 VIS product was transitioned to 4 km resolution 
      on Wednesday, February 19; the GOES-9 VIS product was transitioned to 4 
      km resolution on Wednesday, March 5. This will be the fourth modification 
      made to the contents of the Unidata-Wisconsin datastream as a result of 
      recommendations by the Unidata User Committee at its fall '96 meeting, and 
      their reaffirmation of the transition at their Spring '97 meeting. We are 
      requesting that all sites that have concerns about the change in the imagery 
      products contact the Unidata User's Committee (usercomm@unidata.ucar.edu) 
      and Unidata Support (support@unidata.ucar.edu) as soon as possible. Unidata 
      Unix IDD participants: The larger GOES image products require that you increase 
      your LDM product queue size. We recommend that you increase your queue size 
      by at least 10 MB for every hour's worth of data you retain in your queue 
      to accommodate the six products that are being transitioned from 8 to 4 
      km resolution (GOES-8/9 VIS, IR and H2O) and the CIMSS-produced GOES-8/9 
      sounder products that we will be transmitting as the Research Floater in 
      the near future. Unidata McIDAS-OS2 IDD participants: We are investigating 
      whether the increase in image sizes will require a modification to the IMPORT 
      staging file used by the McIDAS-OS2 decoders. Any site that experiences 
      problems ingesting/decoding the new products should contact Undiata Support 
      for help. Unidata McIDAS Users: The change in the resolution of the GOES-8 
      IR image does NOT require modifications to your file routing table UNLESS 
      you need to reduce the number of products you keep due to disk space limitations. 
      After we finish changing the image products in the Unidata-Wisconsin datastream, 
      will be providing new Function Key Menu templates for McIDAS-OS2 and McIDAS-X 
      that are tailored for the new resolutions. We will announce the availability 
      of the new menu templates when they are ready. Until we release new Fkey 
      templates, you will experience problems with the VIRTual graphic file(s) 
      used as map overlays from the Menu. You can always display images/loops 
      with user-defined loads to get correct map overlays.

Dan Vietor, Purdue,

 Date: Tue, 25 Mar 1997 14:22:25 -0500 (EST) 
      From: Daniel Vietor  Reply-To: devo@cell.atms.purdue.edu 
      To: Unidata Support  Cc: support-datastream@unidata.ucar.edu, 
      Linda Miller , David Knight  
      Subject: Re: 970324: IDD reception of Unidata-Wisconsin datastream On Mon, 
      24 Mar 1997, Unidata Support wrote: > At the end of the recently held User 
      Committee meeting, you and Dave > Knight (SUNY Albany) embarked on an experiment 
      to see whether splitting > the data requested from your upstream feeder 
      would help to mitigate > data reception problems you were having. I have 
      tried a couple of things but with little luck. First I tried splitting the 
      feed with both feeds going to Albany. No luck. Second, I split the feeds, 
      FOS from Albany and McIDAS from Illinois which was a bit better but not 
      by much. Third, I reversed it with no luck. Splitting the feeds does seem 
      to help get the larger images in but it is only marginally better than without 
      splitting. The one thing that is seemingly helping is upping the product 
      queue timeout to 90-120 minutes. I am getting lag times of 4000-4200 from 
      Illinois and 4400-4700 out of Albany. More of the products are getting through 
      but we are still losing about 50% of the various feeds from around 18-22Z. 
      This might all be due to network traffic and uncontrollable but the various 
      tests are not revealing any solutions. ======================================================================== 
      Daniel Vietor INTERNET devo@cell.atms.purdue.edu Dept of Earth & Atmospheric 
      Sciences TITLE Senior Project Specialist Purdue University WXP Developer 
      West Lafayette IN 47907 WXP http://wxp.atms.purdue.edu PH 765-494-3292 FAX 
      765-496-1210 EAS http://meteor.atms.purdue.edu

Peter Schmid

Date: Thu, 17 Apr 1997 12:35:01 +0000 (WET)
From: "Peter M. Schmid" 
To: usercomm@unidata.ucar.edu, mcidas-x@unidata.ucar.edu,
    mcidas-os2@unidata.ucar.edu, Unidata Support

Subject: GOES-8 IR resolution change

I figured I would chime in my thoughts and fears for the increased 
resolution on the satellite images.  The 4Km resolution is great from a 
meteorological standpoint and as far as teaching it significantly helps 
.... BUT ..... there is a defininate problem with the size of the 4Km 
satellite pictures.  They are 2.4 MB in size!  That is a very large file 
for the internet to handle in one large chunck.  I don't care how large a 
bandwidth an individual campus has (as noted by Jim Koermer at PSC with 
2 T1's).  The fact is for most of us the data must travel through major 
internet hubs and go through several peering places in it travel 
through the 'net.  This is where many of the packets are getting 
dropped.  Increasing the bandwidth on the end-user will not totally 
solve this problem.  And as we know the 'net is getting increasingly 
congested.  And it has showed no signs of slowing down.  And ISP's are 
having a very tough time keeping their heads above water with this (some 
are suceeding and some are not:) ).

Maybe there is a way, similar to the broadcast of the HRS data, that we 
can use on these satellite files.  Break them into 4 smaller pieces so
they 
are more likely to get through.  And then post-process them together at 
the end-user.  This, IMO, is a better way of handeling this problem.  It 
is great to have the 4Km stuff but not at the expense of missing data 
during the daytime (prime teaching hours).  Especially now that 
convection time is approaching.  Try teaching convection with out any 
recent satellite images.  It just does not work.

Thanks for listening.

Pete S.

############################################################
# Peter M. Schmid        phone :(802) 626-6328 or 626-6268 #
# Met Lab Coordinator    e-mail:schmidp@apollo.lsc.vsc.edu #
# Lyndon State College   gopher:apollo.lsc.vsc.edu         # 
# Lyndonville, VT 05851  www   :http://apollo.lsc.vsc.edu  #
############################################################

Gilbert Sebenste

 Date: Thu, 17 Apr 1997 08:40:44 -0500 (CDT) From: 
      "Gilbert L. Sebenste"  To: "Peter M. Schmid"  
      Cc: usercomm@unidata.ucar.edu, mcidas-x@unidata.ucar.edu, mcidas-os2@unidata.ucar.edu, 
      Unidata Support  Subject: Re: GOES-8 IR resolution 
      change Just a chime-in from Northern Illinois University... We get all the 
      huge files here just fine. While I completely understand that we need to 
      have the AREA files for teaching, at the same time, the greatly increased 
      resolution helps to teach things that are now visible that were blurry or 
      invisible at 8KM resolution: outflow boundaries, lake breeze fronts, lake 
      effect snow "streaks" in the Great Lakes, and so forth. Previously, we were 
      going to the Internet and UCAR's high-res images to point these features 
      out using Netscape and the WWW. One suggestion was to split the reception 
      of the files into several pieces; any other suggestions would be great, 
      but it would hamper us to take the high resolution images away. Gilbert 
      ***************************************************************************** 
      Gilbert Sebenste, volunteer \/ Assistant McIDAS/DD+ Administrator * /\ _||_ 
      * Northern Illinois University | / \ Internet: sebenste@geog.niu.edu \/ 
      /______\ * Phone: (815)-335-2213 X249 (W) |\_ | _ | * (815)-756-1931 (H) 
      ___________|__________|__|_|_|________________ ********

James Koermer

 Date: Wed, 16 Apr 1997 14:05:59 -0400 From: "James P. Koermer"  
      To: usercomm@unidata.ucar.edu, support@unidata.ucar.edu Cc: ldm-users@unidata.ucar.edu, 
      mcidas-os2@unidata.ucar.edu, mcidas-x@unidata.ucar.edu, joez@psc.plymouth.edu 
      Subject: RE: 970416: 4 km GOES-8 IR product transition to occur on 970423 
      Even though our networking reliability has improved dramatically over the 
      last several weeks and our school has just added a second T-1 line. I want 
      to express my concern over the further transition to 4 km data. With our 
      very good reception of late, the products that we fail to receive most often 
      without question are the 4km VIS area files. The file size is so large that 
      retransmission timeouts are occuring during the peak daytime traffic periods. 
      I'm not getting totally skunked, but I still can count on losing about 4-6 
      of these files each day. Nighttime traffic isn't a concern, but then again 
      most of our instruction is during the daylight hours. My 8KM file reception 
      is nearly 100%. I am not a happy camper with this decision. -- James P. 
      Koermer E-Mail: jamesk@psc.plymouth.edu Professor of Meteorology Office 
      Phone: (603)535-2325 Natural Science Department Office Fax: (603)535-2723 
      Plymouth State College WWW: hhtp://vortex.plymouth.edu/ Plymouth, NH 03264