|
||||||
|
||||||
National Snow & Ice Data Center, Boulder, CO
This document is being continually updated. If you have questions, comments, or suggestions regarding the use of these data or this document, please contact NSIDC User Services via this form.
This file is intended as a quick question and answer reference guide. More detailed information about the data are available from the following data set documents:
There are a number of differences and the relative importance of those differences depends your intended use of the data. We'll summarize here but we recommend you read all the data documentation as well.
Input data: The GISMO is designed to work with gridded input data while the PSQ is designed to work with swath and scene input data. Both GISMO and PSQ produce gridded output data.
Interface: There is a lot of reuse between these two interfaces, and they are designed to accomplish similar tasks, so they look quite similar. There are, however, two main differences.
The GISMO allows searches on only one collection at a time, but it can search on one or more parameters within that collection. The PSQ allows searches on any number of collections and any number of parameters.
The user may not select the grid in the geographic screen of the GISMO. The displayed grid is the grid the selected collection is in. PSQ input data, however, are not already in a grid. Consequently PSQ users can, and should, select their desired output grid in the geographic screen.
Searching: The GISMO only allows searches of one collection at a time and the geographic coverage of every granule in a given gridded data set is the same. Consequently the GISMO only searches by temporal range and parameter. The GISMO does not search by spatial extent since every granule in the collection has the same geographic coveraged.
The PSQ does search by spatial extent since the spatial coverage of the SSM/I swath data and AVHRR scene data varies significantly. The PSQ uses an orbit search algorithm to search for SSM/I swath data and a slightly more restrictive version of that same algorithm to search for AVHRR scene data. These algorithms depend on orbit and sensor characteristics, so the PSQ runs a slightly different search for each data set. Consequently searches on a wide variety of data may take a bit longer.
Output data: The GISMO back end processing simply subsets the data by extracting the users region of interest out of the existing grid. The PSQ back end processing grids the data on the fly. The portion of the swath or scene data that intersects the users region of interest is gridded to the users chosen grid. This makes the PSQ a bit more versatile, but also a bit more CPU intensive.
For more specific information on the differences between the GISMO and the PSQ see questions #7 and #9.
Java applets written in Java2 will not work in most browsers without the plugin, and the PSQ is written in Java2.
Java2 offers additional functionality that is unavailable in Java1. The results panel makes the most use of Java2 and has a number of capabilities that would not have been possible had we used Java1. These include: results can be sorted by any column, columns can be resized, columns can be moved, and a range of rows can be selected by dragging the mouse. Several other features of the PSQ would have been possible using Java1, but would have been more difficult to implement and more awkward to use.
They are on the next page. The small icons above and below the scrollbar on the right hand side of the results screen are the "Page Up" and "Page Down" buttons. The status message at the top of the results panel tells you what page you are on and what subset of the complete result set you are looking at. There is an upper limit of 10,000 granules returned. For more information see the Results section of the help page.
It will. The PSQ interprets this as a search for data from the First Century A.D. Similarly, the date panel will accept three digit years and the PSQ will interpret that as a search for data from the First Millenium. The temporal coverage of the data available from the PSQ starts in late 1987. For more information on the temporal coverage of specific data sets in the PSQ see the data set documentation:
AVHRR Polar 1 km Level 1b Data Set
DMSP SSM/I Brightness Temperatures
The latitude and longitude files are binary arrays of 4-byte integers. Divide the values by 100,000 to get the actual latitude or longitude. For ASCII output that division has already been done.
There are a number of differences and the relative importance of those differences depends your intended use of the data. We'll summarize here but we recommend you read all the data documentation as well.
Resolution:Two standard AVHRR gridded products are available via GISMO: the AVHRR Polar Pathfinder Twice-Daily 1.25 km EASE-Grid Composites and the AVHRR Polar Pathfinder Twice-Daily 5 km EASE-Grid Composites.
The AVHRR Polar 1 km Level 1b Data Set, consisting of 1 km AVHRR Local Area Coverage (LAC) and High Resolution Picture Transmission (HRPT) scenes, are the input data for the 1.25 km product and gridded AVHRR data from PSQ.
AVHRR level 1b 4 km GAC swath data are the input data for the 5 km product. These level 1b data are not available for use with PSQ.
Temporal coverage: The AVHRR Polar 1 km Level 1b Data Set, available for use with PSQ, extends from 1992 through 2003. There are no plans to acquire additional AVHRR data acquired after 2003. Currently, most scenes in the AVHRR Polar 1 km Level 1b Data Set acquired through 2000 are available for PSQ processing. Most of the remaining scenes in the data set through 2003 should be available for PSQ processing before the end of 2005; however, the 1.25 km APP product, which is based on the Level 1b data, extends only through 1998. Numerous gaps exist where Level 1b data are available but are not currently planned to be processed into 1.25 km APP product files. For example, no data for 1996 has been or will be processed. Thus, the temporal coverage for 1.25 km AVHRR data is much better for PSQ than it is for GISMO.
The 4 km AVHRR L1B data set is not available for use with PSQ. The temporal coverage of the 5 km APP data set available via GISMO extends from 1981 through 2000 with very minor gaps.
Spatial coverage: The AVHRR Polar 1 km Level 1b Data Set covers roughly 60% of the Northern Hemisphere AVHRR EASE-Grid and about 30% of the Southern hemisphere AVHRR EASE-Grid. Spatial coverage is limited by the availability of LAC and HRPT data in the polar regions. Thus 1.25 km APP data available through both GISMO and PSQ have limited spatial coverage
The 4 km AVHRR LIB data set covers near 100% of both the northern and southern polar regions. Spatial coverage is comprehensive due to the global availability of GAC data. Thus 5 km data available through GISMO have much greater spatial coverage than the 1.25 km data.
Grid choice: The 1.25 km and 5 km products available in the GISMO are gridded to one of the standard AVHRR EASE-Grids, while the PSQ allows you to choose what grid you would like the data in. The standard grids are among the grid choices in the PSQ so if you have used the standard products in the past you may continue to use the same grid.
Single scenes: The 1.25 km and 5 km products are twice daily composites. Since there may be multiple scenes covering the same area those composites may discard some of the redundant data. The PSQ processes only one scene at a time so all the data are used. However, a particular scene may not necessarily have sufficient coverage of your region of interest.
Regridding: PSQ development reused existing software wherever possible and filled in gaps where necessary. The PSQ backend processing for AVHRR is largely a reuse of the same processing system used to generate the 1.25 km and 5 km products. One drawback of that system is it was built to create only the 1.25 km and 5 km products. Consequently the PSQ backend processing system produces data in the standard AVHRR EASE-Grids and then regrids the data to the users chosen grid. If you are worried about data degradation during this regridding step we recommend requesting your data in one of the standard AVHRR EASE-Grids.
The binary data are in the same format as the 1.25 km product, as discussed in the data set documentation. ASCII output is just the binary data expressed as integers. The data set documentation also discusses the scaling of those integers.
There are a number of differences and the relative importance of those differences depends your intended use of the data. We'll summarize here but we recommend you read all the data documentation as well.
Grid choice: The standard products available via the GISMO or on CD are gridded to one of the standard SSM/I EASE-Grids or Polar Stereographic grids, while the PSQ allows you to choose what grid you would like the data in. The standard grids are among the grid choices in the PSQ so if you have used the standard products in the past you may continue to use the same grid.
Single swath: The standard products are twice daily composites. Since there may be multiple orbits covering the same area those composites may discard some of the redundant data. The PSQ processes only one swath at a time so all the data are used. However, a particular swath may not necessarily have sufficient coverage of your region of interest.
Interpolation: PSQ development reused existing software wherever possible and filled in gaps where necessary. The PSQ backend processing for SSM/I is largely a reuse of the processing system used to generate the standard products. The SSM/I processing system has a fair amount of flexibility built in so the raw data can be gridded directly to the users chosen grid.
The standard SSM/I EASE-Grid Pathfinder products use Backus-Gilbert optimal interpolation (see the data set documentation for more information). The standard SSM/I Polar Stereographic products use drop-in-the-bucket averaging (see the data set documentation for more information). The PSQ uses inverse distance squared interpolation to generate SSM/I gridded data (see the data set documentation for more information).
See Intercomparison of Resampling Methods for SMMR Pathfinder in EASE-Grid Format for a discussion of the differences and the reasons we chose to use the inverse distance squared method in the PSQ. SMMR data will be available in the PSQ in the future.
Binary data are in the same format as the standard product, as discussed in the data set documentation. ASCII output is just the binary data expressed as integers. The data set documentation also discusses the scaling of those integers.
If you have any questions please contact NSIDC User Services.
NSIDC User Services
National Snow and Ice Data Center
CIRES, 449 UCB
University of Colorado
Boulder, CO 80309-0449 USA
phone: +1 303.492.6199
fax: +1 303.492.2468
form: Contact NSIDC User Services
e-mail: nsidc@nsidc.org
