Examining the Discrepancies
In the comparisons of ground to satellite data, there
were some cases where there were time/space matches, but the observations
did not agree. Here is another way to look at the comparison.
| Summary of 99 Ground/Satellite Matches: Comparison of cloud layers seen by satellite vs. ground | ||||
|---|---|---|---|---|
| Clear | Single | Multi | ||
| S | Clear | 14 | 9 | 0 |
| A | Single | 3 | 29 | 3 |
| T | Multi | 3 | 29 | 9 |
In order to better understand why these mismatches occur, each category will be
examined separately.
There are many possible explanations as to why the observations did not match. However, there are some common problems that were found in the comparisons. One problem in matching the ground and satellite data was simply the issue of converting local time to universal time. This was a problem because there are things such as daylight saving time that are sometimes forgotten during the time conversion, resulting in an inaccurate time report. This causes the cloud report to be of little use in satellite comparisons, since it is not actually for the time it says it is.
Another problem possibly encountered on the ground is the horizon. It is possible that landforms or buildings obstruct the field of view, and therefore students are not able to see and report on the entire sky. (Director's Note: This is OK, as we expect such random error effects. If you are in a crowded area, we do recommend observing from the same location each time.)
Related to this problem is the grid system used by the satellite to determine location. The satellite uses a pre-determined one-degree by one-degree grid, regardless of where the school might be. Therefore the satellite data might be for a somewhat different part of the sky than the ground observers' report, and the cloud observations might not match due to this problem.
A final common issue is the problem of one thick cloud layer obscuring the view of another cloud layer, and therefore only one of the layers is reported. For example, if fog were to obstruct the view of the ground observers, they would not report high cirrus clouds, simply because they could not see them. (Director's Note: Again, this problem is expected. The S'COOL data may be used to help us quantify how often it occurs.)
These general issues explain some discrepancies in the data comparisons. Now let's look at each type of mismatch individually.
| Summary of 99 Ground/Satellite Matches: Comparison of cloud layers seen by satellite vs. ground | ||||
|---|---|---|---|---|
| Clear | Single | Multi | ||
| S | Clear | Match (14) | Ground Single Layer/Satellite Clear (9) | 0 |
| A | Single | Ground Clear/Satellite Single Layer (3) | Match (29) | Ground Multiple Layers/Satellite Single Layer (3) |
| T | Multi | Ground Clear/ Satellite Multiple Layers (3) | Ground Single Layer/Satellite Multiple Layer (29) | Match (9) |
