Difference between revisions of "Process representation across tools"
| Line 181: | Line 181: | ||
== Surface runoff to channel network == | == Surface runoff to channel network == | ||
| − | '''Note on surface runoff vs ponding and surface storage:''' In the Pitman models, all water not calculated to infiltrate in a timestep, which is a month, is assumed to be runoff. For models with shorter timesteps, water reaching the land surface which does not infiltrate into soil in a timestep can become surface runoff or remain as surface ponding/surface storage. In subsequent model timesteps, water still on the land surface can evaporate, infiltrate, and/or become surface runoff in the next timestep. Some models have relatively short timesteps (subdaily, daily) compared to the rate at which water would move all the way across a modelled land unit (which could be large, rough, and/or flat, slowing the flow rate). This is why some water will be 'surface storage' in one timestep and then 'surface runoff' in the next timestep. MIKE-SHE can also consider that land surfaces can be very rough, or have dips that trap water, and so some amount surface water will not be able to run off at all. This water will both evaporate and infiltrate over time. | + | '''Note on surface runoff vs ponding and surface storage:''' In the Pitman models, all water not calculated to infiltrate in a timestep, which is a month, is assumed to be runoff. For models with shorter timesteps, water reaching the land surface which does not infiltrate into soil in a timestep can become surface runoff or remain as surface ponding/surface storage. In subsequent model timesteps, water still on the land surface can evaporate, infiltrate, and/or become surface runoff in the next timestep. Some models have relatively short timesteps (subdaily, daily) compared to the rate at which water would move all the way across a modelled land unit (which could be large, rough, and/or flat, slowing the flow rate). This is why some water will be 'surface storage' in one timestep and then 'surface runoff' in the next timestep. MIKE-SHE can also consider that land surfaces can be very rough, or have dips that trap water, and so some amount surface water will not be able to run off at all (detention storage). This water will both evaporate and infiltrate over time. |
</br> ACRU4 differs notably, assuming the water that does not infiltrate into soil accounts for both surface runoff and interflow runoff. | </br> ACRU4 differs notably, assuming the water that does not infiltrate into soil accounts for both surface runoff and interflow runoff. | ||
{| class="wikitable" | {| class="wikitable" | ||
Revision as of 15:27, 26 November 2023
PAGE IN PROGRESS - MORE COMING SOON!
building process representation overview tables + links to more specific pages about individual processes
Canopy interception and evaporation (vs throughfall)
| Algorithm description |
WRSM-Pitman | SPATSIM-Pitman | ACRU4 | SWAT2012 | MIKE-SHE, semi-distributed & fully distributed |
|---|---|---|---|---|---|
Algorithm inputs (input data, parameters, model-calculated states/storages) |
|
|
|
|
|
| Function type |
exponential & threshold | exponential & threshold | threshold | threshold | |
| Thresholds |
yes:
|
yes:
|
yes:
|
yes:
|
Infiltration into soil moisture (vs surface runoff or surface ponding)
Note: This excludes the case of a modelling unit (HRU, grid cell, area within a subcatchment, etc.) is designated as impervious. In this case, rain inputs would become surface runoff, potentially with a portion staying behind as surface storage/ponding if the area has attenuation specified (to represent roughness and flatness).
| Algorithm description |
WRSM-Pitman | SPATSIM-Pitman | ACRU4 | SWAT2012 | MIKE-SHE, semi-distributed, more conceptual |
MIKE-SHE, fully-distributed, more physical |
|---|---|---|---|---|---|---|
Algorithm inputs (input data, parameters, model-calculated states/storages) |
|
|
|
|
|
|
| Function type |
non-linear & threshold | non-linear & threshold | power & threshold | power & threshold | linear (rate) & threshold | non-linear & threshold |
| Thresholds |
yes:
|
yes:
|
yes:
|
yes:
|
yes:
|
yes:
|
Surface runoff to channel network
Note on surface runoff vs ponding and surface storage: In the Pitman models, all water not calculated to infiltrate in a timestep, which is a month, is assumed to be runoff. For models with shorter timesteps, water reaching the land surface which does not infiltrate into soil in a timestep can become surface runoff or remain as surface ponding/surface storage. In subsequent model timesteps, water still on the land surface can evaporate, infiltrate, and/or become surface runoff in the next timestep. Some models have relatively short timesteps (subdaily, daily) compared to the rate at which water would move all the way across a modelled land unit (which could be large, rough, and/or flat, slowing the flow rate). This is why some water will be 'surface storage' in one timestep and then 'surface runoff' in the next timestep. MIKE-SHE can also consider that land surfaces can be very rough, or have dips that trap water, and so some amount surface water will not be able to run off at all (detention storage). This water will both evaporate and infiltrate over time.
ACRU4 differs notably, assuming the water that does not infiltrate into soil accounts for both surface runoff and interflow runoff.
| Algorithm description |
WRSM-Pitman | SPATSIM-Pitman | ACRU4 | SWAT2012 | MIKE-SHE, semi-distributed, more conceptual |
MIKE-SHE, fully-distributed, more physical |
|---|---|---|---|---|---|---|
Algorithm inputs (input data, parameters, model-calculated states/storages) |
*Month timestep: all water not infiltrating becomes runoff
|
*Month timestep: all water not infiltrating becomes runoff
|
*Lag coefficient separates "quickflow"~surface runoff from "delayed-flow"~interflow. Quickflow portion reaches the channel on the same day generated (rain day). The rest is lagged over subsequent days |
|
|
|
| Function type |
(no transformation) | (no transformation) | (no transformation) | non-linear | non-linear & threshold | non-linear & threshold |
| Thresholds |
no*
|
no*
|
no*
|
no*
|
yes:
|
yes:
|
Evapotranspiration (ET) from soil moisture (SM)
Note: More coverage of evapotranspiration related terminology and inputs across different tools can be found here
| Algorithm description |
WRSM-Pitman | SPATSIM-Pitman | ACRU4 | SWAT2012 | MIKE-SHE, semi-distributed, more conceptual |
MIKE-SHE, fully-distributed, more physical |
|---|---|---|---|---|---|---|
Algorithm inputs (input data, parameters, model-calculated states/storages) |
*determines linear decline of ET with soil moisture decline
|
*determines linear decline of ET with soil moisture decline
|
*ET assumed to also decline if soil gets close to saturation, waterlogging, unless wetland plants
|
|
|
|
| Function type |
linear & threshold | linear & threshold | multi-part linear & threshold | non-linear & threshold | multi-part linear & threshold | non-linear & threshold |
| Thresholds |
yes:
|
yes:
|
yes:
|
yes:
yes:
|
yes:
|
yes:
|
Evapotranspiration (ET) from groundwater (GW)
Note: More coverage of evapotranspiration related terminology and inputs across different tools can be found here
| Algorithm description |
WRSM-Pitman (Sami GW) |
SPATSIM-Pitman (Hughes GW) |
ACRU4 | SWAT2012 | MIKE-SHE, semi-distributed, more conceptual |
MIKE-SHE, fully-distributed, more physical |
|---|---|---|---|---|---|---|
Algorithm inputs (input data, parameters, model-calculated states/storages) |
|
|
|
+ all inputs for ET from soil
|
+ all inputs for ET from soil
|
+ vegetation property inputs for ET from soil
|
| Function type |
non-linear & threshold | non-linear & threshold | non-linear & threshold | non-linear & threshold | non-linear & threshold | |
| Thresholds |
yes:
|
yes:
|
yes:
|
yes:
|
yes:
|