The input data will be converted to a single output interval and data type, which is selected by the user. The method for converting from input to output time interval is dependent on the input data type, the output data type, and a comparison of the input time interval and output time interval. See the rules below.
Transformation Rules:
Input Data Type
Output Data Type
Is input interval > output interval?
Operation
Description
PER-CUM
PER-CUM
Yes
Dis-aggregation
Use transformation pattern to convert to the output time interval, by applying the factors to the input image. By default, uniform weights are used, which conserves mass.
PER-CUM
PER-CUM
No
Aggregation
Sum up all input images within the output interval. Apply the defined weight (weight is 1 by default, so mass is conserved).
PER-AVER
PER-AVER, INST-VAL
Yes
Dis-aggregation
Applies input image values to all output intervals within that input interval.
PER-AVER
PER-AVER
No
Aggregation
Compute the average of all input images within the output interval.
PER-AVER
INST-VAL
No
Selection
Select the values from the input image in which the output date is coincident in time.
INST-VAL
PER-AVER
Yes
Interpolation
Interpolate to the center of the output interval, and assume that is the average over the interval.
INST-VAL
PER-AVER
No
Aggregation
Compute the average of all input images within the output interval, but correcting for the timing of the input image within the output interval. For example, if the input interval is 1Hour, but the output interval is 2Hour, there will be 3 input images to average. The first and last image will be given a weight of .25, and the middle image will be given a weight of 0.5.
INST-VAL
INST-VAL
Yes or No
Interpolation
Perform interpolation to get the appropriate time step.
INST-VAL, PER-AVER
PER-MIN
No
Selection
Select the minimum value from the input images that fall within the output interval.
INST-VAL, PER-AVER
PER-MAX
No
Selection
Select the maximum value from the input images that fall within the output interval.