Currently input data files must be given with the same timestep as the reading model component, but many externally-provided files may be on a different timestep. This stipulation does not appear to apply to inwards data, which can be (dis)aggregated as needed to enable compatibility between components.
A simple first improvement would be to allow the user to specify an aggregation/disaggregation method for a given variable. For instance, daily rainfall might be uniformly disaggregated to hourly or summed to monthly as appropriate. The user would need to choose sensible methods - UniFHy should not make assumptions about the method to use, and some more complex schema would be too hard to implement (e.g. sinusoidal varying hourly temperature derived from daily min/max).
Currently input data files must be given with the same timestep as the reading model component, but many externally-provided files may be on a different timestep. This stipulation does not appear to apply to inwards data, which can be (dis)aggregated as needed to enable compatibility between components.
A simple first improvement would be to allow the user to specify an aggregation/disaggregation method for a given variable. For instance, daily rainfall might be uniformly disaggregated to hourly or summed to monthly as appropriate. The user would need to choose sensible methods - UniFHy should not make assumptions about the method to use, and some more complex schema would be too hard to implement (e.g. sinusoidal varying hourly temperature derived from daily min/max).
Suggested methods:
Disaggregation:
Aggregation: