A quick (and hopefully painless) introduction to the DMSP SSIES database


              We know you wish to get to the data as quickly as you can, but please take a moment to go through this list of things you need to know before you use these data.  We also urge you to read the frequently asked questions section (FAQ) for a more complete explanation about the database and the data, as well as the other pages off of the main page where we provide a more detailed background on just about everything.




  The last point here is the most important: how reliable are these data?  In an ideal world every pass in this database would have been reviewed by one of us and then either approved (making it level 2 quality data) or reprocessed until it was of satisfactory quality.  However, there are a total of twelve satellite-years of data here already (and ultimately the complete dataset will have over 50 satellite-years of data) and there are only two of us (Drs. Hairston and Coley) working on this project part-time.  There is simply no way we can do that given the limits on our time.  So we have done the next best thing: we have developed an automatic routine which evaluates the quality of each 4 second datum from the RPA and IDM as described above.  Although this routine is automatic, we have designed it to err on the side of caution, thus if there is the slightest question about a datum it is flagged as caution (flag=2), poor (flag=3), or undetermined (flag=4).  Essentially all of the data marked as good (flag=1) are most certainly good and can be used with a high degree of confidence.  As you will see after examining several of these orbital plots, once something happens to the data, such as the light ion percentage starts to increase or the Senpot device fails in sunlight (as happened on F14 in the 2000-2002 period), the quality of the data goes down quickly and significantly, and there is simply not much we can do to salvage these data by reprocessing.  Thus, while we will be refining this quality algorithm over the next several years, it is doubtful that more than a few percent of the data currently labeled as "caution / poor / undetermined" will ever be reclassified as good.



Back to the main page.