2.2 Dealing with missing data |
|||||||||||||||
Images: atsr199701.gif atsr199801.gif Description Download images (195 K)
Useful information: |
To create a useful histogram we need to prevent missing data (land and cloud pixels) from being included in the plot. This can be done by designating the value of 255 as a null value. You may remember that you have done this already, when you opened the HDF file in section 1.8. Designating a null valueA null value is basically the numerical value (or range of values) assigned to data points where valid data do not exist. To set the null value in an image, you first need to open the Redisplay dialog. You may do this in different ways:
From the menu bar:
The effect of 'Applying Stretches'When you created the previous two histograms you made sure that the 'Apply Stretches' box on the 'New' dialog was unchecked. To see why this is important, create a third histogram based on the whole image, but this time make sure you check the Apply Stretches box before clicking OK to create the histogram. Study Histograms 2 and 3 and use your observations to answer Question 2 below. |
||||||||||||||
Answers:
Back up to: |
Before you continue, close the two histograms that do not represent the underlying SST data correctly (cfr. your answer to to Question 2c). |
||||||||||||||
|