National Environmental Satellite, Data, and Information Service Home Page National Oceanic and Atmospheric Administration Home Page National Climatic Data Center Home Page Department of Commerce Home Page
NOAA Logo, National  Environmental Satellite, Data, and Information Service. National Climatic Data Center, U.S. Department of Commerce

NOAA National Operational Model Archive & Distribution System

Miscellaneous Information

Table of Contents

Building Dataset Collections
Working with Timeseries Data
Creating Plots with GrADS
Scheduled Downtime Notification

Top of Page Building Dataset Collections

On the Model Data Access page you are given a table that contains one entry for each possible model-grid grid combination. When you choose either the "plot" or "ftp" functions, you first have to build a "dataset collection" for the model-grid you have chosen. Both functions take you to the same form. You build a dataset collection by selecting a time period, cycles, and forecast hours for your model-grid. Your collection contains files from all model runs initialized within the time period you specify.

After you build your dataset collecion, you are given a page where you can review the collection. You can choose to revise your collection or continue to "plot" or "ftp" data. Both the "plot" and "ftp" function work only on the set of files in your collection. You will only have access to data contained within your dataset collection.

A dataset collection is a logical grouping of the files you have selected. When you build a collection, you are given a unique 4 digit order number. This order number is displayed on the page where you can review your collection. This order number is used to identify files if you later choose to put any data on our anonymous ftp server for retrieval. If you choose to revise your collection, you keep the same order number.


Top of Page Working with Timeseries Data

The NOMADS Web Interface allows users to easily constuct a timeseries across all the files in a collection by using the template function in GrADS. We dynamically build a ctl file for a timeseries of just the analysis fields (i.e., forecast hour 000) from all the model runs in your collection and then use the ctl file and gribmap to generate the necessary index file. This timeseries is built from all the "_000.grb" forecast hour files in your collection. Since this ctl file spans multiple cycles, we use "hh" in the filename to identify the file as follows:

meso-eta_218_20030501_hh00_000.ctl <-- aggregate of all 000 forecast hours

Use this ctl file when you want to construct any kind of timeseries plots from analysis fields.

The first page of the web interface offers you and option to create the above file. The checkbox
"Generate control file for analysis fields over all cycles"
below the list of cycles and forecast hours toggles this feature.

Note: Turning this feature on requires gribmap to be run on the new template control file. Thus, on larger datasets this can take a very long time. The checkbox described in the above paragraph is unchecked by default for datasets we do not recommend its usage


Top of Page Creating Plots with GrADS

After you build and review your dataset collection, you can use GrADS to generate plots. Press "Plot Data". First you select a ctl file from one of the model runs you have included in your collection. Then you select a type of plot: map, timeseries, lat,lon vs time, or animation. If you want to construct a timeseries of analysis fields across multiple runs, then select the "hh" ctl file. See the "Working with Timeseries Data" section for details. Then you select a parameter and level, and specify other attributes. We use GrADS to generate the requested plot. You can download a postscript version of the plot or the GrADS meta file.


Top of Page Scheduled Downtime Notification

Until we can get a redundant load-balance front-end server system in place, there will be times when we need to take down the server for maintenance. All scheduled server and data outages will be posted below the data access page table of contents at least 24 hours before the scheduled downtime. If you notice a dataset you need is unavailable, please check this notification panel before contacting us. Also, notices regarding new datasets, dataset changes, and dataset terminations will also be posted to this panel.