3. Our webservice requires the creation of a URL
query based on user inputs (such as coordinates of
bounding rectangle, time period of interest and
parameter of interest) which will then be used to
request data from the website. Could you provide
information on how to convert geographical
coordinates
(in decimal degrees) and time values (MM/DD/YYYY
hh:mm) into the array indices shown in the website
above? e.g. time[0:1:0] y[0:1:0] x[0:1:0]
this is not trivial, as it depends on the specific
file. there are routines in the netcdf-java 2.2
library that can do it. Are you using java?
Hi Ernest:
Ernest To wrote:
John,
My name is Ernest To and I am a student of Dr.
Maidment's. Tim Whiteaker and I are trying to
write a
webservice that would download total precipitation
data for the continental US at resolution of 20km
from
unidata website for a given time and place. We
have
so far been experimenting with the following link:
Can we ask you the following questions?
1. What is a .grib1 or a .grib2 file? Is it in
netcdf format?
no, grib1 and grib2 are WMO formats for gridded data
2. We noticed that in the link above (OPENDAP )
there
are options for getting a binary file or an ASCII
file. Are those files in netcdf and cdl formats
respectively?
no, they are in opendap specific formats. normally
you would use the opendap client library (C++ or
Java) to access data through opendap, rather than
downloading files.
3. Our webservice requires the creation of a URL
query based on user inputs (such as coordinates of
bounding rectangle, time period of interest and
parameter of interest) which will then be used to
request data from the website. Could you provide
information on how to convert geographical
coordinates
(in decimal degrees) and time values (MM/DD/YYYY
hh:mm) into the array indices shown in the website
above? e.g. time[0:1:0] y[0:1:0] x[0:1:0]
this is not trivial, as it depends on the specific
file. there are routines in the netcdf-java 2.2
library that can do it. Are you using java?
what file formats can you handle? what does your web
service do?
Thank you for your kind attention. Please let
me
know if anything is unclear. I can be reached
either
by e-mail: address@hidden or by phone:
512-508-4834.
Thanks.
Ernest To
PhD student
Center for Research in Water Resources
University of Texas at Austin
512-508-4834
--- John Caron <address@hidden> wrote:
Hi David, Ernest:
Have a look at the motherlode catalog:
http://motherlode.ucar.edu:8080/thredds/
we were guessing you will want the NAM 21-km
dataset?
If you want to use WCS protocol, you can get back
geotiff or netcdf-CF. Otherwise you could use
Opendap.
There's a few other wrinkles concerning time
dimension, we can go over.
David Maidment wrote:
Ernest:
Please see the message below from Mohan
Ramamurthy
at Unidata. Can you
please work with John Caron
to see how we can ingest a field using web
services. I'd like to stick
to something 2D on the land
surface like precipitation, evaporation or soil
moisture. I believe
those are all in the eta model
output. Likely what we want to do is to give a
lat-long box, a time
point and a variable, and ask
for a field over the region of the lat-long box
for that variable and
time point, and it might occur
that the field actually covers several forecast
intervals, eg. 3 hour
forecast times for a day. I'm
not sure in that case if we should treat each
time
slice as one field in
the web services ingest, but
lets look into what is involved and then figure
out what is best.
John -- I'd like to see what can be done with a
WCS service if possible.
Thanks for your help.
David
-----Original Message-----
From: Mohan Ramamurthy [mailto:address@hidden]
Sent: Friday, March 10, 2006 5:40 PM
To: David Maidment
Cc: John Caron; Mohan Ramamurthy
Subject: Web service to Operational Eta model
output
David,
I just spoke to John Caron
(address@hidden), our lead developer
on THREDDS, re. our conversation this afternoon
to
create a web service
to access fields from the operational model
output
in the IDD stream.
John will be happy to work the person on your to
make it happen. It
looks like many of the pieces are already in
place
to enable a web
service for accessing, subsetting, and moving
individual fields, because
the data is served via OPeNDAP. For example,
here
is the OPeNDAP URL
for today's Eta output: