Dear IPCC analyst,

Although the IPCC publication deadline has passed, we plan to continue support 
of the multi-model dataset at PCMDI which most of you have accessed.  We 
expect the current database to be the basis for continued scientific investigation.

We are writing to ask you for input to to assist us in preparing an improved data 
service for the next round of simulations (likely to become available a few years 
from now).  There is an expectation that different parts of the future multi-model 
database will be archived at multiple modeling centers and a few data centers 
around the world, and the data will be accessible through a common portal.  The 
user will be able to search and retrieve the data as if it were all at a single location.

As we plan for this, we would like your input concerning some specific items 
below and, more generally, your thoughts concerning what your needs will be.  
Do not feel obliged to answer every questions; any input you provide will be helpful.

========================================================================================
Please return the completed survey to: williams13@llnl.gov.  
Your input is needed by September 11th to be most useful. 
========================================================================================

1. First, please rate the following items, according to the priority you think they 
deserve (high, medium, or low) :

a) Improved search capability.

b) Improved publication of errata.

c) Improved notice of new data availability.

We are aware that the registration procedure was confusing for some of you, 
and we will work to simplify it.  If you have suggestions as to what features 
should be included in any of the above, please provide them here:



2. To what extent did you use the data portal (https://esg.llnl.gov:8443/index.jsp),
and to what extent did you obtain data via ftp?



3. Would extracting a subset of the data in a file be really useful? (e.g., only 
a certain region, only a few time-slices, only a single layer)?  What about 
aggregating data from several files into a single file?



4. Would you have preferred to obtain data that had to some extent already 
been processed on the server side?  (e.g., climatology computed for some 
standard period or a zonal mean).  If so, specify what type of calculations.



5.  What dissimilarities in the way the output from different models was 
stored caused you the most headaches, and how important would it be to 
impose more uniformity?



6.   Please share with us any negative or positive experiences you had 
dealing with the database at PCMDI.  In particular, what is the single 
thing you would most like to see fixed or improved?




Finally, we are working with our colleagues at several other climate 
research centers to develop software capabilities that help address 
the longer term needs of our community.  Please think expansively 
as you provide guidance concerning the following:

a) What problems in obtaining and analyzing data do you foresee that 
a more powerful portal or integrated network for sharing and analyzing 
climate data could help solve?




b) Are there other tools, services, or capabilities related to serving, 
sharing and analyzing climate data that you would find useful if they 
could be integrated into a data portal system?




c) Do you have any other comments as to what a future data portal 
system should offer? 




========================================================================================

Thank you very much for your input.  Your suggestions will help 
improve our service to you.

Sincerely yours,
Karl Taylor and Dean Williams
PCMDI