HDF Scaling

The general name of the HDF scaling operation is "data scaling". Data scaling refers to the process of converting bit data values to geophysical values using a scaling factor.

In the past, many remote sensing bit data values were provided as 8-bit data, meaning that the data range had 256 values, commonly 1-256, less commonly -128 to +128. More recently, as data storage capabilities have increased, bit data could be 16-bit (65,536 values) or even 32-bit (4,294,967,295 values). 8, 16, and 32 refer to the exponential powers of of 2, i.e. 8-bit data has 28 (or 256) values.

The data scaling operation takes the scaling factor in the HDF data file and uses it to convert the bit value to a geophysical data value. The scaling factor could be simply a multiplier or divisor; more commonly, the scaling factor is an expression, such as a linear expression Y = aX + b, where X is the bit value, Y is the geophysical data value, and a and b are scaling factors. It is also possible to have a logarithmic scaling expression, of the form Y = 10(aX + b), or other mathematical expressions.

This Technical Summary page is a draft version -- for information on this operation while the documentation is being refined, please send a message to GES DISC Help.