HRU/Observation indexing.
The simplest CRHM model has one set of observations. That is one value of each of the following observations; t, rh, u, and p (or ppt) for every timestep of the model run This is very limiting when the area being modelled is large and diverse in elevation and exposure. CRHM is capable of indexing each HRU to a different observation element using the parameter HRU_OBS. This allows every HRU to use a different set of observations. To further expand the flexibility, the observations are divided into the five groups described below.
The simplest would be if the HRU and the observation were aligned in sequence.
| HRU | 1 | 2 | 3 | 4, ... |
| Observation | 1 | 2 | 3 | 4 |
The following illustrates a random arrangement.
| HRU | 1 | 2 | 3 | 4, ... |
| Observation | 1 | 7 | 5 | 2 |
If the indexing table requests a value that does not exist, the highest value defined is used. In the above, if there are only 4 elements in the temperature dataset, the HRUs requesting the 5th and 7th array would be truncated to 4. The highest possible element index.
Five HRU/Observation arrays.
The observations are grouped into five categories according to how the field data is collected or the probable database source.
- t, rh, ea.
- u.
- p and ppt.
- Q or other radiation components.
- miscellaneous category to be used as required.
| HRU | 1 | 2 | 3 | 4, ... |
| Observations t, rh, ea.[1] | 1 | 7 | 5 | 2 |
| Observation u. [2] | 1 | 2 | 2 | 1 |
| Observation p, ppt. [3] | 1 | 1 | 1 | 1 |
| Observation Q. [4] | 2 | 2 | 1 | 1 |
| Observation misc. [5] | 1 | 1 | 1 | 1 |
It should be noted that for this system to be most useful, the observations for the five categories should be in five separate files. This is optional as long as the observations are addressed accordingly.