Model units & connections

This page describes how modelling tools allow users to discretise catchments into modelled units in order to represent and calculate hydrological processes. Differentiating the catchment into separate components or units - such as subcatchments, patches of similar land cover, or soil layers with distinct properties - allows hydrological processes to be modelled by algorithms that have been developed for that scale and type of unit.

Each modelling tool has a unique way of describing a catchment in terms of:

• The types surface and subsurface model units that can be included
• The connections that can be defined between these different units
• The process algorithms that are used to calculate inflows, storage, and outflows for each unit

This page summarises the basic unit types and connection options (i.e. flow paths) available across the tools. Sets of tables further describe how the distribution of land cover types, soil types, aquifer types, river channels, and water bodies in a catchment can be represented using each tool. These also cover options for vertical distribution of soil, sediment, and rock types. The potential linkages or routing of water between units and layers are highlighted.

Process algorithms, the rules and equations that govern flows in and out of units and layers, are described on a processes page here. Although covered on separate pages, the approach to spatial unit and vertical layer discretisation and the process algorithms used are inextricably linked.

This page also describes some general implications of the structural differences across the tools. Their importance can be case specific. The implications of the differences across these tools in some particular model application contexts are described in sections on specific use cases, here.

The material on this page may be more depth than you are looking for. If all you need at the moment is an overview of the structures and capabilities of the different modelling tools, that summary can be found on the capabilities overview page, here

Importance of understanding structure options

To set up a model of a specific catchment that is relevant to our understanding of it's processes (conceptual model) entails attention to modelling units and how they are, or are not, connected to one another. A flow path that we think is important in the catchment, and/or is important for the particular modelling project, may not be directly represented by certain modelling tools.

For example, not all tools can include channel transmission loss from rivers, and those that do include it differ in where the water goes after leaving the model channel.

When a flow path or process that we believe to be important in our use case cannot be explicitly included using a given modelling tool, we can:

• find a way to represent it implicitly, through purposeful parameter adjustment or other work-around (e.g. increasing certain storages, synthetic input or abstraction routes, etc), or
• choose to use another modelling tool.

Understanding how modelled units, subcatchments, HRUs, channels, reservoirs, etc, can interact is also important when deciding how to discretise the catchment. Strategies may need to differ when using different tools.

For example, SWAT2012 and SPATSIM can both include irrigation fed from storage dams; however, the two tools have different limitations in terms of where a dam can sit in a model subcatchment and in relation to the irrigated area. In SWAT2012, a 'reservoir' can be an irrigation source. There can be one reservoir per subcatchment and it is located specifically on the main channel at the subcatchment outlet. A reservoir can irrigate an area outside of it's subcatchment. In contrast, SPATSIM-Pitman has a similar 'reservoir unit', but only models irrigation from 'dams', which are different. A 'dam' is located internally in a subcatchment, not on the main channel, and is fed by a proportion of the runoff of the incremental subcatchment only. A dam can only irrigate area within the same subcatchment. These two approaches could lead to quite different subcatchment delineation decisions to represent the same thing!

In general, model unit and connection differences impact:

• the processes that can or cannot be explicitly represented in a model (depending on unit types and scales, some processes will be represented in the link between two or more model units)
• the algorithms that are appropriate for explicit processes, and therefore the meanings of the parameters and the appropriate values for these parameters

NB: Different tools have some model input parameters with the same or similar names, based on shared overarching concepts (e.g. saturated soil moisture storage). However, if the parameters are used in different algorithms and/or applied at different spatial and temporal scales across tools, the values that are appropriate for an area will differ by tool!

• delineation decisions for subcatchments and other units, because these decisions will have tool-specific impacts on:

representing spatial climate variability,

the number of land cover types that can be included in the model

connectivity of surface and subsurface flows in the landscape,

how waterbodies and irrigated areas can be fed

the locations within the catchment that a model can produce different predicted outputs for

• if and how the location of a land cover type within a subcatchment (e.g. upland vs riparian) is represented and how this will impact predicted processes.

Back to top of page

Basic approaches for representing catchments in the different modelling tools

The schematic diagrams and tables below describe the main discretisation approaches used in each tool in terms of breaking up catchments into units and specifying linkages.
For MIKE-SHE, because the tool offers many options, two generalised set-up approaches have been described (more background on these is given here).

Schematic diagrams of spatial discretisation

Subcatchment delineation Demonstration of: (a) a catchment area delineated into subcatchments and (b) the flow network connections of these subcatchments when used as model units. This approach is used to some degree in all the modelling tools covered (except MIKE-SHE when fully distributed). Subcatchments 4,5,6,8 are not headwaters and their river channels receive flow from upstream subcatchments. Figures from: Schulze, R.E. (1995). Hydrology and Agrohydrology: A Text to Accompany the ACRU 3.00 Agrohydrological Modelling System (Water Research Commission).	Subcatchments as units for process modelling Diagram illustrating the approach taken in SPATSIM-Pitman of using the subcatchment as the primary unit for representing processes. Flow through the subcat can be conceptualised as flow through a generalised hillslope, which provides a means of modelling groundwater flow and exchange with the river channel. Special subareas (e.g. irrigated crops) impact the subcat's infiltration, surface runoff, ET, and recharge. Interflow and aquifer exchange with the channel are calculated based on the resulting subcatchment-scale storage. WRSM-Pitman & SWAT follow a related approach for groundwater in that aquifers are modelled at the subcatchment scale. Figures from: Hughes, D.A. (2004). Incorporating groundwater recharge and discharge functions into an existing monthly rainfall–runoff model. Hydrological Sciences Journal, 49.	Modular network of modelling units: Example WRSM-Pitman model network diagram (catchment of the Churchill Dam, quaternaries K90A&B), demonstrating a catchment composed of modules linked by routes. Subcatchment boundaries are not set-up in the tool; however, sets of linked modules are subcats in effect. (Also, though shown as separate units, each 'irrigation module' is modelled as part of a specified 'runoff module' for groundwater modelling.) ACRU4 uses a similar modular network approach, although ACRU units are grouped into subcatchments. Figure from: Bailey, A.K., and Pitman, W.V. (2015). Water Resources of South Africa 2012 Study (WR2012) (Water Research Commission).
Networks of modelling units within subcatchments Demonstration ACRU4 model network diagram showing HRUs (as coloured squares), river channel, and dam units grouped within subcatchments. Climate can be input at the subcatchment scale and certain relationships may only be allowed between units in the same subcatchment (e.g. routing HRU subsurface flow to a special riparian HRU). The figure also shows that withdrawals and external inputs can be included for rivers and dams. Figure from: Clark, D.J., Smithers, J.C., Thornton-Dibb, S.L.C., and Lutchminarian, A. (2012). ACRU 4 User Manual: User Interface & Tutorials in Volume 3 of Deployment, Maintenance, & Further Development of SPATSIM-HDSF. (Water Research Commission).		HRUs as units for process modelling: Schematic diagram of process representation for an HRU (hydrological response unit) in ACRU4. Each HRU includes a groundwater/baseflow store. SWAT2012 also uses HRUs to model surface and shallow subsurface processes, but models groundwater at the subcatchment scale. Figure modified from: Schulze, R.E. (1995). Hydrology and Agrohydrology: A Text to Accompany the ACRU 3.00 Agrohydrological Modelling System (Water Research Commission).
Routing runoff from units in parallel or in series Demonstration of subcatchment delineation into topographic units, which could serve as HRUs or process modelling zones, and related routing options. Runoff from each unit could be routed directly to the stream in parallel, or routed from one unit to the next in a downslope series. Series routing allows processes (infiltration, recharge, saturation seepage, etc) to occur along the path depending on states of downslope units (dry/saturated, flat/steep, rough/smooth, etc). It adds computational complexity. Different approaches can be more appropriate, or more important, in different settings, spatial scales, and for different runoff components (i.e. surface flow, interflow, groundwater flow). SPATSIM subareas in subcats modify subcat-scale runoff; no separate routing needed. WRSM's afforestation & alien veg modules are similar (act inside 'runoff modules'), but irrigation module surface runoff and interflow are routed in parallel to the linked 'runoff module'. ACRU4 routes all HRU 'quickflow' runoff in parallel and can optionally route 'baseflow' runoff into the soil of a specified 'riparian HRU'. SWAT2012 routes all HRU surface and 'lateral' (interflow) runoff in parallel, while groundwater is modelled at the subcat scale. MIKE-SHE grid cells are connected in series. If 'overland flow zones' are used, surface runoff can be routed in series or in parallel. If interflow and baseflow linear reservoirs are used, interflow is routed in series and groundwater in parallel (if multiple aquifers used).
3D grid cells as units for process modelling Illustration of the approach used in MIKE-SHE for modelling processes using a 3D grid, i.e. grid cells on the surface (2D) with columns of material below that are broken into layers. (For those familiar with HRUs, each cell can be seen as similar to an HRU connected in series to its neighbors.) In the diagram part of the unsaturated zone is shown as transparent so that the gridded water table surface beneath can be more easily visualised. Figure source: DHI (2019). MIKE SHE Manual, Volume 1: User Guide, MIKE 2017 (Danish Hydrologic Institute).		Hybrid approach - different unit types for modelling different processes Illustration of the more lumped, conceptual approach for representing interflow and groundwater flow with 'linear reservoir' units in MIKE-SHE. Other processes can still be modelled by grid cell. Interflow reservoirs have spatial extents and each receives percolating water from the grid cells that overly it. Similarly if multiple different mapped baseflow reservoirs (distinct aquifers) are included, these are recharged by overlying interflow units. SWAT2012 also uses a hybrid approach in that surface flow and 'lateral' flow (interflow) is modelled at the HRU scale, while the aquifer spans the subcatchment an is recharged by all the overlying HRUs. Figure source: DHI (2019). MIKE SHE Manual, Volume 2: Reference Guide, MIKE 2017 (Danish Hydrologic Institute).

Back to top of page

Model structure overviews by tool

NB: For MIKE-SHE many structural options are available. Two generalised set-up approaches are used in these comparisons to represent the spectrum. These two are described on another page, here.
In addition, the different tools use slightly different terms for the same or similar concepts. This is documented on the terminology page, here. In general these sections use terms as they are used in the tool being described.

Level / Element	WRSM-Pitman (Sami GW)	SPATSIM-Pitman (Hughes GW)	ACRU4	SWAT2012	MIKE-SHE, semi-distributed, more conceptual	MIKE-SHE, fully-distributed, more physical
Catchments & subcatchments (subcats)	WRSM catchments are composed of ‘modules’ linked in a user-specified network. Module types: runoff modules, special area modules (see below) channel modules reservoir modules Subcatchments are not explicitly input, but runoff modules and sets of linked modules act as subcats.	SPATSIM catchments are composed of subcatchments. Subcats are linked in spatially determined flow network (map input).	ACRU4 catchments are composed of subcatchments Subcats are linked in a user-specified flow network.	SWAT catchments are composed of subcatchments Subcats are linked in a spatially determined flow network (delineated by SWAT from DEM input, or given as a map input)	MIKE catchments can be composed of subcatchments Subcats are linked in spatially determined flow network (map input).	MIKE catchments can be composed of 3D grid units. The surface is uniformly sized grid cells. Each cell has a column beneath: input layers of soil, sediment, rock. No subcat boundaries are input; grid cells have elevation (DEM input) and water elevation on the surface and in the 3D grid drive flow.
Units within subcatchments	A WRSM ‘runoff module’ with no linked special areas functions as a subcat. If special area modules are linked, then a set of modules functions as a subcat.   Special area modules that can be added are: afforestation (tree plantation), alien vegetation, irrigated crops mines These appear as separate units, but function as sub-areas inside a specified linked runoff module. Runoff modules can also have internally specified subareas: impervious area riparian area (ET from GW).	SPATSIM subcatchments all contain a river channel and may have specified special sub-areas. Subcats can optionally include either a wetland or a reservoir/lake on the channel at the subcat outlet. A special sub-area can be defined for any, or all, of the following in a subcat: impervious area riparian area (ET from GW) area with a higher ET land cover (forest, tree plantation, alien veg) crop irrigated from dams crop irrigated from river small dams internal to subcat (lumped) Impervious, high ET cover, and irrigation areas do not have defined spatial locations within the subcat.	ACRU4 subcatchments are composed of one or more HRUs. Non-headwater subcats must contain a river channel. Subcats can optionally include additional channels and reservoir units. Units are linked in a user-specified network. There are special HRUs for: impervious areas, irrigated areas, riparian areas wetlands. For most HRUs, all runoff is routed directly to a channel, reservoir, or subcat outlet node (in parallel). Optional exceptions: For ‘riparian HRUs’, ‘baseflow’ output from other HRUs can be routed to the riparian HRU soil. For ‘adjunct-impervious HRUs’, runoff is routed to the surface of another HRU.	SWAT subcatchments are composed of: HRUs (many) shallow aquifer unit, deep aquifer unit, tributary channel unit main river channel unit. Subcats can include an internal pond or wetland and a reservoir/lake on the channel at the subcat outlet. HRU surface & ‘lateral’ runoff output is routed directly to the subcat channel (in parallel). This passes via the tributary channel, where delays & infiltration can occur, to the main channel. Groundwater is modelled at the subcat scale. Aquifers are recharged by all subcat HRUs. Shallow aquifer outflow goes to the subcat main channel.	MIKE-SHE subcatchments are composed of: grid cells (veg & soil) ‘overland flow zones’, ‘interflow reservoirs’ ‘baseflow reservoirs’ river channel links Surface runoff is lumped for grid cells within an overland flow zone & routed across zones in a downslope series to the channel (potential for infiltration on route). Interflow reservoirs receive percolation from overlying grid cells. Interflow is laterally routed through a downslope series of reservoirs to the channel (potential for loss to recharge on route). Baselfow reservoirs are recharged by overlying interflow reservoirs and outflow to the channel in parallel.	Using fully distributed MIKE-SHE, no subcat boundaries are input. Relative water surface or head elevations on the surface and in the 3D grid units drives flow exchange between units.   Surface and subsurface properties for the grid cells & layers are input for different zones (map input). Different sets of zone polygons can be used for the different types of properties (e.g. vegetation types, soil types, aquifer extents); they do not need to be aligned with one another.
River channels	WRSM channels are modules in the network with multiple inflow and outflow links possible. Channel modules can have an associated wetland storage/area.	SPATSIM channels are units linked to subcats (one per subcat). They receive the local subcat runoff & upstream subcat channel outflows.	ACRU4 channels are units in a subcat network of units. (necessary in non-headwater subcats). They receive runoff from linked HRUs, reservoirs, & upstream subcat outflows.	SWAT channels are units linked to subcats (one main channel per subcat). They receive the local subcat runoff & upstream subcat channel outflows.	MIKE channels are composed of spatially explicit reaches between node points. Reaches with nodes mapped inside a subcat exchange water with the subcat.	MIKE channels are composed of spatially explicit reaches between node points. Reaches can exchange water with grid cells that border them.

At a basic level, these modelling tools are similar in some notable ways:

• Catchments divided into subcatchments: All the tools can represent a larger catchment as an accumulative flow network of smaller subcatchments. In this way they are all ‘semi-distributed’ to a certain extent (with MIKE-SHE also offering ‘fully distributed’ – fully gridded – representation as an option).

• Same major vertical layers: All the tools represent processes for the same broad set of vertical layers - vegetation canopy and land surface, soils/unsaturated zone, and aquifer materials – which are then linked to a channel network. This is not the case for all catchment models. For example, the original Pitman model (Pitman 1973) did not separate soil and aquifer processes, but instead used a lumped sub-surface storage unit to handle both.

• Multiple land cover types per subcatchment: All the tools allow different land cover types to be explicitly represented across different subcatchments and within them. At a minimum, within each subcatchment all the tools can at least explicitly represent:

a dominant generalised vegetation type per subcatchment (i.e. local indigenous veld type, non-irrigated grazing lands)

a separate tree-dominated vegetation type per subcatchment (e.g. forest, commercial tree plantations, invasive alien tree stands)

irrigated crops

impervious cover (urban area, bare rock).  

• Reservoirs/dams/waterbodies: All tools can include reservoirs/dams/waterbodies fed by the channel network.    

Back to top of page

Discretisation and process representation

The scale of discretisation of the landscape influences which hydrological processes are individually represented and the algorithms and the timesteps that are appropriate (discussed in more detail here). In general, for larger spatial/vertical units, longer timesteps, more lumped process representation, and different property parameters are applicable compared to modelling with smaller units. This has to do with how long it’s likely to take for water to move through a modelled unit. For example, it may take a week for interflow to move through the 10 km long hillslope of subcatchment, but a day to move through the 1 km slope length of a patch of grassland within that subcatchment. Speed also varies by process and with conditions (e.g. dry and wet) which also needs some consideration in the algorithms used with a uniform calculation timesteps.

Differences in algorithms and parameters with spatial scale also have to do with the fact that what may be modelled as one process at a larger spatial scale can be the result of several different processes occurring at smaller scales. For example, rain falling on a steep, rocky cliff may form surface runoff. If this surface runoff then flows across an area of permeable and unsaturated soil before reaching a stream channel, some may infiltrate and not reach the stream as surface flow. If one models a subcatchment where this is happening within a single spatial unit, the equation and parameters for estimating ‘surface runoff generation’ would account for the net outcome of the surface runoff that actually reaches the stream, i.e. the combined impact of the cliff and the permeable toe-slope on surface runoff reaching the channel. If modelling the cliff and the toe-slope as separate units, surface runoff on the cliff unit could be calculated, and then surface flow and infiltration on the toe-slope unit could be calculated as a separate set of processes, to get to the amount of surface runoff reaching the channel. 

Timesteps & spatial units for different processes

Units	WRSM-Pitman (Sami GW)	SPATSIM-Pitman (Hughes GW)	ACRU4	SWAT2012	MIKE-SHE, semi-distributed, more conceptual	MIKE-SHE, fully-distributed, more physical
Timesteps	Monthly* A daily version has been developed. Limited use to-date	Monthly* A daily version has been developed. Limited use to-date	Daily	Daily, subdaily	Daily, subdaily* Timesteps are dynamic, vary by process & unit size. Outputs saved for the same selected step.	Daily, subdaily* Timesteps are dynamic, vary by process & unit size. Outputs saved for the same selected step.
Spatial units for:
Climate input	Modules (all)	Subcatchments	Subcatchments or HRUs* *HRU-scale input is labor intensive	Subcatchments	Grid cells or zones	Grid cells or zones
Surface & shallow subsurface processes	Runoff modules + special area modules	Subcatchments (+ internal special sub-areas)	HRUs	HRUs	Grid cells + Overland flow zones + Interflow reservoir zones	Grid cells
Groundwater processes	Runoff modules	Subcatchments	HRUs	Subcatchments	Baseflow reservoir zones	Grid cells

Summary of structural differences & implications

Despite their high-level similarities, there are numerous differences in the way that these tools allow users to discretise and represent various components of catchments in terms of scales, unit types, and connections. These differences have implications for what a model built with that tool can do (see capabilities overview table). They are also important for the discretisation and parameterisation choices a user must make.

The specific implications of the differences across tools, and their real importance for a modelling project, will differ across use-cases. These will be dependent on the combination of the type of catchment, the changes that are to be modelled, the data available, and the types and scales of model outputs that are needed. This makes it difficult to generalise about the implications or relative importance of a particular difference. Nevertheless, the tables below demonstrate some potential implications of some highlighted structural differences as examples:

• CLIMATE: Spatial units for which climate inputs are specified
• LAND COVER: Units used to represent different land cover types & their connectivity
• AQUIFERS: Units used to represent aquifers & their connectivity
• RIVER CHANNEL NETWORK: Channel units & links to landscape units

Further implications of tool differences in some particular applied contexts are described in a section on specific use-cases, here.  

CLIMATE: Spatial units for which climate inputs are specified
Why it can be important & when	Determines if/how spatial variation in climate over the catchment can be represented in the model. Likely more important to model outcomes for catchments with: larger gradient magnitudes across the catchment area (e.g. larger catchment size, more mountainous terrain) generally drier areas in which the higher rainfall parts are frequently the only places where thresholds for runoff production are reached (i.e. if more averaged rainfall were applied for a larger area, then no runoff would be modelled) NB: Potential advantages of including more spatial climate variability in a model also depend on the level of data available about the spatial distribution of rainfall and PET.
Alternative approaches	Subcatchment	HRU / module / waterbody	Individual grid cells or independent zones (not tied to the boundaries of other inputs like land cover, subcatchments, etc)
Tools using approach	SPATSIM-Pitman WRSM-Pitman (uniform within ‘runoff module’ + associated treed modules) SWAT2012 ACRU4 (most user-friendly way to input)	ACRU4* (possible, but very labour intensive) WRSM-Pitman - limited* (irrigated areas, reservoirs, wetlands can have own climate)	MIKE-SHE (with both semi- & fully-distributed options)
Potential implications of approach	Areas with important differences in climate need to be delineated as separate subcatchments to include this. This may entail compromises between directly representing climate variability and directly representing flow pathways that are effectively broken by subcatchment boundaries. Implications vary by tool: in some there is no GW flow between subcats and areas feeding dams, wetlands, riparian areas & irrigation may need to be in the same subcat (see aquifer implications and specific structure tables below for details)	Can include spatial variability in climate within a subcatchment. Avoids potential compromises in connectivity that could be imposed by subcat boundaries (see notes to left). HRU or module delineations may be altered to include climate gradients: i.e. breaking up a broad land cover type into multiple units that receive different climate inputs. In ACRU4 & WRSM, setting up HRUs and modules is a many-step, manual process. The added labour and opportunity for user error would depend on the number of units involved.	Spatial variability in climate can be included without the need to compromise on representing hydrological connectivity and with relatively little added effort (see notes to left). Climate inputs and land cover, soil, and other properties are each specified using independently delineated zones and then separately assigned to each grid cell by the modelling tool. Delineating climate zones (deciding number & spatial extent) and preparing zone data is a step in all approaches. This approach avoids some decision-making about where/how to compromise (see notes to left).

LAND COVER: Units used to represent different land cover types & their connectivity
Why it can be important & when	Influences the number of different land cover types that can be explicitly represented and what types of land cover changes can be modelled. The importance of this will depend on the purpose of modelling project and the diversity of cover types within the modelled area. NB: Potential advantages of being able to explicitly represent more different land cover types in a model will also depend on the property information available to reliably parameterise them as being different from one another.   Influences if/how the specific location of a land cover within a landscape is considered when modelling its processes (e.g. upland vs riparian). NB: Relevant when modelling at spatial scales that can capture position differences vs larger scales that must average over them Likely more important in more water limited areas and with diverse topography: position is a more significant determinant of water access.
Alternative approaches	Each subcatchment has one ‘main’ land cover type and additional types are represented as proportions of the subcatchment area. Each additional cover type is represented with different type-specific algorithms (e.g. tree plantation, irrigated crops)	Land cover properties are assigned to HRUs within subcatchments, or to individual grid cells. These units (HRUs, grid cells) all use the same basic process algorithms with respect to land cover - differences between cover types come from the property parameter inputs.
Tools using approach	SPATSIM-Pitman WRSM-Pitman	ACRU4 SWAT2012 MIKE-SHE
Potential implications of approach	Fewer different cover types can be explicitly included vs other approaches. Particularly limits the number of non-irrigated, non-afforested types (e.g. veld types, grazing land, levels of degradation). In theory, more types can be included using more (smaller) subcats. This becomes unwieldly to set-up, limits aspects of connectivity (see above), and algorithms may not apply to very small subcats. SPATSIM allows fewer types over all (see table), but has some flexibilities. It allows two types of non-irrigated cover per subcat: a portion is assigned a higher ET rate. This rate is flexible, so could be a second veld type, plantation, invasives, or a lumped combination of these. WRSM has modules for tree plantations and invasive alien trees. Both can be added in one subcat, but their algorithms are specific to those covers. The riparian area within a ‘runoff module’ can be given it’s own ET pan-factors, allowing it to be another cover type flexibly defined. This could cause ‘double counting’ with the treed-area modules that effectively act on the full runoff module. (See riparian areas page LINK for more.)   A more detailed cover composition can be represented implicitly, with parameters for broader model types calculated from the local mix of more detailed types. With conceptual parameters, this may not be straightforward. Most cover types don’t have an explicit spatial location within a subcat. The riparian zone of a WRSM runoff module is an exception. The SPATSIM riparian area does not have its own cover type. (See riparian areas section for more.)	No tool-imposed limit on the number of different cover types that can be explicitly included in a model.   Limitations will come from the information available to reliably parameterise types as different from one another as well as from the practicality of setting up many types in the tool (see interface section) Linkages allowed between units (HRUs, grid cells) determine if/how the location of a cover type within a subcat is explicitly considered. If each unit’s runoff is routed directly to a channel in parallel to others, without interacting, relative position is not represented.     Position is explicit in MIKE-SHE: flows are routed into/onto neighbouring units or zones. This is more limited in ACRU4 and SWAT2012, which have mostly parallel routing, but there are ways in which location (i.e. riparian vs upland) can be represented for certain HRUs, hence cover types. In ACRU4, special riparian HRUs receive ‘baseflow’ runoff from linked upslope HRUs. Special riparian and wetland HRUs are also linked channels and can receive overflow. In SWAT2012,the level of access to groundwater for ET can be set by HRU. This can be set higher for some HRUs to differentiate riparian vs uplands. Aquifers are subcat scale, so uplands can recharge the store used by riparian. Special ‘depression’ waterbodies can also be set-up on specified HRUs, which receive routed runoff from linked upslope HRUs.

AQUIFERS: Units used to represent aquifers & their connectivity
Why it can be important & when	Determines if/how spatial variability of aquifer material (sediment deposits, geological formations, layers) within a subcatchment can be explicitly represented (if information on this is available). This will be more important when there are large known differences at the subcat scale (depends on choice of subcat sizes). Determines the scale at which groundwater storage, and potentially water table depth, can be output by the model. Determines where within in a model catchment groundwater (GW) can outflow to a channel. If there is no subsurface GW flow between modelled units, then any recharge within a unit that is thought to contribute to outflow of the larger catchment will need to emerge as baseflow from that same unit in the model. In reality there may be subsurface flow with the GW emerging further down. Having the GW emerge further upstream can impact other modelled processes linked to channel flows at different locations (overbank flow, transmission loss, waterbody storage, irrigation). Influences the groundwater access of vegetation in different parts of the catchment, which can impact modelled seasonal and total ET and the modelled impact of vegetation cover changes. (Linked to model representation of the location of a land cover type – see above). Likely more important to model outcomes for: catchments where GW contributes more substantially to the overall hydrograph, cases where disconnected units (e.g. subcats) are small in comparison to connectivity of aquifers important to the hydrograph (delineation accounting for other factors, i.e. climate & land cover) areas where GW is an important water source for vegetation (particularly GW recharged by a large surrounding area)   modelling projects focused on low flows and/or flows at points within the catchment
Alternative approaches	Subcatchment-scale aquifers No GW flow between subcats	Subcatchment-scale aquifers GW can flow between subcats	HRU-scale aquifer units (or other units smaller than subcatchments) No GW flow between aquifer units or subcats	3D grid of connected volume units All units can interact
Tools using approach	SWAT2012 MIKE-SHE, semi-distributed, linear reservoir GW option* (can have a subcat-scale unit)	SPATSIM-Pitman WRSM-Pitman	ACRU4* (partial exception: riparian HRU) MIKE-SHE, semi-distributed, linear reservoir GW option* (can have multiple units in a subcat; these don’t interact)	MIKE-SHE, fully distributed, finite difference GW option
Potential implications of approach	Subcat aquifer parameters reflect average, or net, storage and flow properties of aquifer material in the subcat. Subcat scale GW model outputs. Within a subcat: model can allow vegetation in one area to access GW that was recharged in another location in the subcat. Between subcats: If its important to for veg in a downstream subcat to access GW that was recharged in an upstream subcat, work-arounds are needed. Water to feed this ET would need to leave the upstream subcat as channel flow and then become accessible downstream, e.g. through channel transmission loss. Alternatively revise the subcat delineation. To model total catchment outflow, some GW may need to enter the channel network in upstream model subcats that in reality flows in the subsurface and emerges further down (downstream subcats in model). This could influence aquifer parameterisation choices (i.e. vs field data on properties). If having more water in the channel higher up in the catchment has significant impacts on other processes (upstream dams, etc), adaptations may need to be made to these and/or the subcat delineation revisited.	Subcat aquifer parameters reflect average, or net, storage and flow properties of aquifer material in the subcat. Subcat scale GW model outputs.   Models can allow vegetation in one area to access GW that was recharged in another location in the same subcat or another subcat (only applies to the specified ‘riparian zones’ in SPATSIM-Pitman & WRSM-Pitman subcats/runoff-modules). Avoids potential subcat delineation and aquifer parameterisation compromises or water routing work-arounds (see notes to left)	Aquifer parameters can differ across units within a subcat to represent spatial distribution of different aquifer materials (e.g. fractured rock in uplands, alluvium over rock in lowlands). GW model outputs can be obtained for different areas within a subcat. Separated model GW stores that do not interact implies separate aquifers outflowing to channel network independently. This may not be the case in reality. This would impact aquifer parameter choices (i.e. vs. field property data). Within a subcat: Vegetation can only access GW stores that are recharged in the same spatial unit (HRU in ACRU4, baseflow reservoir extent in MIKE-SHE). Lowlands cannot access GW recharged in highlands in the model if these are separate HRUs or zones. This could result in unrealistically low ET. ACRU4 & MIKE-SHE have optional routines to overcome this. ACRU4’s special riparian HRU can receive ‘baseflow’ output from linked upland HRUs (routed to riparian soil).  In MIKE-SHE, water from ‘baseflow reservoirs’ can be routed to the soil of a riparian zone to make up a deficit (feed ET).   Implications of also not modelling GW flow between subcatchments are the same as those described to the left.	Aquifer parameters can differ across units within a subcat to represent spatial distribution of different aquifer materials (e.g. fractured rock in uplands, alluvium over rock in lowlands). GW model outputs can be obtained for specific grid cells in the model as needed. Because there can be flow between all aquifer units (determined by relative head and conductivity between neighbouring cells) and no subcat boundaries limiting subsurface connectivity, vegetation in lowlands would be able to access GW recharged elsewhere in the model when conditions allow. The approach can have high computational demand, so model runs are longer (runtimes depend on computing power).

RIVER CHANNEL NETWORK: Channel units & links to landscape units
Why it can be important & when	Determines the locations within the catchment where modelled streamflow outputs will be available.   This can be important if there are monitoring points or other points of interest where these outputs would assist, i.e. water supply withdrawal points, areas for flood assessment, aquatic habitat assessesment. Influences where reservoirs, waterbodies, wetlands that are fed by the channel network can be located with respect to model subcatchments and what areas feed them. Determines the potential for representing spatial variability in channel properties, where these are understood and thought to be important Flow from channel units to landscape unit surface & subsurface can allow for water exchanges between delineated subcatchments (see notes above). Channel processes that are not explicitly included (e.g. transmission losses, overbank flooding) will be implicitly represented, likely through parameterisation, to generate net aquifer outflow and net surface runoff.  This will be more important in catchments where channel transmission loss and/or overbank flooding are frequent and make a notable impact on the flows.
Alternative approaches	One channel reach unit per subcatchment Two-way channel-landscape sub-surface exchange Limited or no overbank flooding	Flexible number and arrangement of channel reach units, manually connected to HRUs, modules, reservoirs Two-way channel-landscape sub-surface exchange Limited or no overbank flooding	Channel reach units spatially linked to landscape grid cells and have explicit elevation All units can interact, surface and subsurface
Tools using approach	SWAT2012 SPATSIM-Pitman WRSM-Pitman* (some channel exchange processes are calculated within ‘runoff-modules’)	ACRU4 WRSM-Pitman* (separate channel modules in network)	MIKE-SHE
Potential implications of approach	To get model output streamflow for a point, it needs to be the outlet of a subcatchment. Reservoirs on the channel network will be at the outlets of subcatchments (not in WRSM), so also influence subcat delineation (see specific waterbody table below). Subcat delineation also has implications for subsurface flow connectivity and other aspects of representation (see above) – various compromises may be needed in the delineation. SPATSIM & SWAT2012 partially address the restriction in reservoir location by including other waterbody unit types that can be internal to subcats, fed by proportions of subcat runoff. The set-ups of these types have other restrictions (See waterbody table below). Subcat-scale channel & subcat-scale aquifer representation facilitates calculation of exchange (GW to channel, channel to GW, “transmission loss”) considering storage and properties of both. This is done in SPATSIM and in WRSM within a run-off module. In SWAT2012, channel transmission loss goes to a separate bank storage unit, not the subcat-aquifer. SPATSIM & WRSM include overbank flooding only in wetland modules. SWAT2012 does not include overbank flooding onto land units (wetland module not on the channel). Channel transmission loss in SWAT2012 & SPATSIM and overbank flooding to wetland units provide means by which downstream areas can access water that entered the channel network from upstream units.	Flexibility in where streamflow outputs can be obtained and where reservoirs/waterbodies and transfers can be located. This can reduce some delineation related compromises. In ACRU4, channel units are fed by linked HRUs. This will impact the delineation of HRUs (e.g. an area of a single cover type in a subcat may be split into multiple HRUs to feed into different points in the channel network, above or below a reservoir, etc). The manual process of setting up HRUs and establishing routing connections individually in ACRU4 means that taking advantage of this flexibility comes with significant additional labour. In WRSM, user-specified portions of the outflow of a single ‘runoff module’ can be routed to multiple channel and reservoir modules. These channels and reservoirs are then effectively inside the subcat represented by the runoff-module. This reduces the number of individual runoff-modules that need to be set-up.   Conversely all runoff from a ‘runoff module’ can be routed to one channel module and/or multiple runoff modules can contribute to a channel module. This allows modules with different land cover compositions to contribute as different points, more like ACRU4 HRUs.       ACRU4 does not explicitly represent channel transmission loss. WRSM calculates this within runoff modules. Additional ‘bedloss’ can be specified for channel modules, but this is removed from the model, not added to an aquifer, not available for ET. Implication: water entering the channel network from upstream landscape units cannot be accessed for ET in downstream landscape units via channel transmission loss. (see section on channel transmission loss LINK) ACRU4 & WRSM include overbank flooding only in their wetland units/HRUs and ACRU4 riparian HRU.  This is a means by which downstream areas can access water that entered the channel network from upstream units.	Model streamflow outputs can be obtained for any point in the channel network. Reservoirs and water bodies can be added at any location. Because these have explicit locations in the model grid, the approach of lumping many small farm dams into a single unit can difficult to operationalise depending on the drainage network, etc.   Using fully distributed options: Explicit location and elevation in the model grid allows calculation of exchanges between channel reaches and their neighbouring cells based on the water levels in each and conductivity. The explicit depth of the channel impacts how much groundwater enters it. Bank topography can limit surface flow entering the channel. Overbank flooding and channel transmission losses occur when levels allow. The approach comes with notable practical implications.  It requires careful vetting of landscape topography data and the channel cross section inputs to make sure these correspond so that flow exchanges are realistically modelled. The method has high computational demand, so model runs are longer (runtimes depend on computing power). Using the more lumped surface and groundwater options: Surface and GW flows into the channel aren’t calculated using the explicit elevation of the channel. Channel transmission loss and overbank flooding can still be included using different algorithms. These provide means by which downstream areas can access water that entered the channel network from upstream units.

Summary comparison of WRSM and SPATSIM-Pitman

WRSM and SPATSIM-Pitman are more similar than they may appear from the structural overview, sharing subcatchment structure & core algorithms

The WRSM-Pitman approach to model structure, with a flexible network of modules, makes it appear to be very different to SPATSIM-Pitman, which uses a clearly defined subcatchment as the basic unit of calculation for most processes. However, looking at the process algorithms for the different 'modules' in WRSM-Pitman makes it clearer that the two tools actually have relatively similar, mostly subcatchment-based, approaches.

The WRSM 'runoff module', on its own with no linked special area modules, functions like a basic subcatchment of SPATSIM-Pitman, with no defined sub-areas of treed cover or irrigated crops. In this case, both tools use the same subcatchment scale, monthly time-step process algorithms of the original Pitman model (Pitman 1973) adjusted to include a separate aquifer store.

The WRSM modules for representing commercial tree plantations and irrigated crops visually appear as separate in the model interface and network diagrams (which are generated by users, not by the software tool), but actually they function as parts of a 'runoff module'/subcatchment area. The linkage to a run-off module is established in their set-up specifications in the software. The process algorithms for these areas differ across the tools, but in both cases the treed cover modules act to increase the ET losses of the subcatchment/'runoff-module'. The resulting runoff (surface runoff, interflow, aquifer flow to channel) is generated at the subcatchment scale. For irrigated crop areas, net runoff is tied to processes calculated in the linked subcatchment in both cases as well. This means that the WRSM treed area and irrigated crop modules function more like sub-areas of subcatchments in SPATSIM than like HRUs in ACRU4, in which more processes are calculated for the HRU unit alone.

Aside from process algorithm differences, the main structural differences across the two tools in terms of modelled units 'inside' subcatchments are:

• the numbers of different special area types than can be included 'in' a subcatchment (i.e. tied to a 'runoff-module' in WRSM),
• the flexibility and arrangement of water sources that can be used for irrigation, and
• flexibility in where reservoirs, withdrawals, and transfers can be placed and how they are fed.

In terms the basic model subcatchment, there is also a potentially important difference in the connection to the channel network between the tools. This is relevant to their representation of groundwater-surface water interactions. Both tools calculate aquifer-river channel exchange for the basic subcatchment unit. This is calculated within a 'runoff module' in WRSM, before the water is routed to a connected channel module. Channel modules receive the net runoff after this exchange, and do not interact further with the runoff module/subcatchment. In a sense, there is an implicit 'tributary channel' within a WRSM runoff module for which certain processes are represented, and the channel modules are a main channel network that is handled differently. In SPATSIM-Pitman, each subcatchment has a channel element that is tied to it. The channels of downstream subcatchments receive channel flow from those upstream and their channel-aquifer exchange is calculated using this accumulated channel flow. In WRSM-Pitman, this exchange is only calculated for the runoff produced in the run-off module, the incremental subcatchment, and channel flow from upstream subcatchments is not involved. The overall impact of this different on catchment-scale modelling outcomes would vary depending on how important this exchange for a given area.

Specific units & connections for different model components

Land cover: spatial distribution & type limitations

Factor	WRSM-Pitman (Sami GW)	SPATSIM-Pitman (Hughes GW)	ACRU4	SWAT2012	MIKE-SHE, semi-distributed, more conceptual	MIKE-SHE, fully-distributed, more physical
Spatial units for land cover type/property distribution	Modules & sub-areas within them	Subcatchments & sub-areas within them	HRUs	HRUs	Cover type polygons applied to grid cells	Cover type polygons applied to grid cells
	Extents of cover types are either defined in the: ‘runoff module’ set-up special area modules (which are tied to, and act as sub-areas of, a ‘runoff module’) Each runoff module can have different general cover specifications & linked special areas.	Extents of cover types are defined as special sub-areas within a subcat. Each subcat can have different general cover & sub-areas specifications.	Each HRU has its own cover & vegetation parameters.	Each HRU has cover & vegetation parameters. HRUs are assigned cover types using an input cover type map. Properties are input by type.	Each grid cell is assigned a cover type using an input cover type map. Properties are input by type.	Each grid cell is assigned a cover type using an input cover type map. Properties are input by type.
Limitations to types & number of types	Yes	Yes	No	No	No	No
	Types that can be represented: General vegetation (runoff module parameters) Impervious area (runoff module parameters) Riparian zone vegetation (runoff module parameters) Afforestation (special module, 1 per runoff module) Invasive alien vegetation (special module, 1 per runoff module) Irrigated crops (special module, multiple per runoff module) Mines (special area module)	Types that can be represented: General subcat vegetation Impervious area Higher ET vegetation (forest, tree plantation, alien veg) Irrigated crops Only one area (one type) of each of these broad classes can be represented within a given subcat.	No limit on number of HRUs (hence cover types) per subcat.	No limit on number of HRUs or cover types per subcat.	No limit to number of cover types.	No limit to number of cover types.

Soils & unsaturated zone (UZ) material: spatial distribution, vertical layers, connections

Factor	WRSM-Pitman (Sami GW)	SPATSIM-Pitman (Hughes GW)	ACRU4	SWAT2012	MIKE-SHE, semi-distributed, more conceptual	MIKE-SHE, fully-distributed, more physical
Spatial units for distribution of soil types / properties	Runoff modules & irrigation modules	Subcatchments	HRUs	HRUs	Soil type polygons applied to grid cells & ‘Interflow reservoir’ extent polygons	Soil type polygons applied to grid cells
	Each ‘runoff module’ can have its own soil moisture store & ‘percolation storage’ properties. Afforestation & alien veg modules use the soil of the runoff module they are linked to / part of. Irrigation modules have their own soil properties.	Each subcat can have its own soil moisture store properties.	Each HRU can have its own soil properties	Each HRU can have its own soil properties. HRUs are assigned soil types using an input soil type map. Properties are input by soil type. Individual HRU soil properties can by modified.	Each grid cell is assigned a soil type using an input soil type map.   Interflow reservoir ‘type’ extents are input as a map (boundaries do not need to align with soil types or subcatchments)	Each grid cell is assigned a soil type using an input soil type map.
Vertical layers in soil or UZ profile	2 Layers	1 Layer	2 Layers	11 Layers (max)	2(3) Layers*	Unlimited Layers
	A runoff modules has 2 ‘UZ’ layers: Soil moisture storage Percolation lag storage	A subcat has a single soil moisture storage unit.	An HRU has 2 soil layers: A horizon   B horizon Both in root zone.	A soil type can have up to 10 layers (user input layers). Profile can extend below root zone. Each HRU also has a vadose zone (lag storage) ‘layer’ below the soil profile	A soil type has vertically uniform parameters (no layers), but each model grid cell has 2 computational UZ layers: Upper layer (root zone) Lower layer (below roots) Interflow reservoirs are storage units fed by all overlying grid cells	A soil type can have an unlimited number of layers (user input layers). Profile can extend below root zone.
Surface runoff routing (w.r.t. spatial units)	Parallel*	(n/a)	Parallel	Parallel	Series or parallel	Series
	Parallel for runoff modules & irrigation modules. Other special area modules modify runoff module flow	Subcatchment scale	Parallel from HRUs in subcat	Parallel from HRUs in subcat	Across a series of mapped flow zones within a subcat OR each zone in parallel	Across grid cells based on elevation.
Interflow routing (w.r.t. spatial units)	Parallel*	(n/a)	Parallel	Parallel	Series	Series
	Parallel for runoff modules & irrigation modules Other special area modules modify runoff module flow	Subcatchment scale	Parallel from HRUs in subcat	Parallel from HRUs in subcat	Through a series of interflow reservoirs in a subcat	Through grid cells based on head.
Capillary rise: 2-way connection to aquifer	Yes*	Yes*	No*	Yes	Yes*	Yes
	Only in riparian zone sub-area of runoff module; represented as ET deficit met by GW	Only in riparian zone sub-area of subcat; represented as ET deficit met by GW	Water in the baseflow store of an ACRU4 HRU cannot move into the soil profile or be used for ET in the same HRU. BUT when upland HRU baseflow is routed to riparian HRU soil this is similar to GW access in riparian zones in subcat scale models....	Represented as ET deficit met by GW; access can be set by HRU	Only for grid cells in riparian zone (lowest interflow zone); represented as ET deficit met by GW

Aquifers: spatial distribution, vertical layers, connections

Factor	WRSM-Pitman (Sami GW)	SPATSIM-Pitman (Hughes GW)	ACRU4	SWAT2012	MIKE-SHE, semi-distributed, more conceptual	MIKE-SHE, fully-distributed, more physical
Spatial units for distribution of aquifer types / properties	Runoff modules	Subcatchments	HRUs	Subcatchments/HRUs*	‘Baseflow reservoir’ extent polygons	Layer & lense polygons applied to grid cells
	Each runoff module can have its own aquifer properties.	Each subcatchment module can have its own aquifer properties. There is a spatial division into two slope sections, riparian & upslope, for process calculation, but these don’t have separate property parameters.	Each HRU can have its own ‘baseflow storage’ properties.	Aquifer GW storage & outflow is calculated per subcat, but properties can be input per HRU to model capillary rise	‘Baseflow reservoir’ type (aquifer type) spatial extents are input as a map (boundaries do not need to align with other inputs)	Each grid cell has a layered profile of aquifer material types based on the layers and lenses it overlies.   Property ‘layers’ cover entire the model domain, can have spatially variable thickness (grid input). ‘Lenses’ occur in certain areas (map input) also with variable thickness.
Vertical layers within aquifers	1 Layer	1 Layer	1 Layer	2 Layers	2 Layers	Unlimited Layers
	A runoff module has one aquifer unit	A subcat has one aquifer store. There a two linked horizontal sub sections, but no vertical divisions.	An HRU has one ‘baseflow’ storage.	A subcat has 2 aquifer units: shallow aquifer (outflow to channel) deep aquifer (no outflow)	A mapped ‘baseflow reservoir’ type can have 2 internal storage units with different parameters.	No limit on the aquifer material layers & lenses that can be input. Calculation grid layers can be set differently to the aquifer property layers. In this case thickness-averaged parameters are assigned.
GW flow routing between spatial units	Yes	Yes	(Limited)	No	No	Yes
	GW flow between subcats	GW flow between subcats	No GW flow between subcats Upslope HRU baseflow can be routed to riparian HRU soil within a subcat No other HRU GW exchanges included	No GW flow between subcats	No GW flow between subcats No GW flow between ‘baseflow reservoirs’ (aquifer types) within subcat	GW flow between grid cells
GW flow routing vertical aquifer units/layers	(n/a – single unit)	(n/a – single unit)	(n/a – single unit)	No	No	Yes
				Shallow & deep aquifers each receive a portion of total recharge and do not interact (more like units than vertical layers)	The 2 units in a mapped baseflow reservoir type each receive a portion of total recharge and do not interact (more like units than vertical layers)	GW can flow between vertical layers based on head and conductivity
GW flow out of model domain (catchment)	Yes	Yes	No	Yes*	No*	Yes
				Only deep aquifer can have GW flow the model. (Deep aquifer does not feed channel, can be pumped)	Recharge can be allocated to ‘dead storage’ which has a similar impact.	GW outflow boundary condition must be set-up
GW abstraction included	Yes	Yes	No	Yes	Yes	Yes
Number of abstraction points allowed	1 per runoff module	2 total per subcat: 1 per subcat aquifer section: 1 upslope & 1 riparian	(n/a)	2 total per subcat: 1 per subcat aquifer unit: 1 shallow & 1 deep	Flexible number, location, & abstraction depth	Flexible number, location, & abstraction depth
GW abstraction routing options	Removed from model	Removed from model	(n/a)	Applied as irrigation, or Removed from model	Applied as irrigation, or Removed from model	Applied as irrigation, or Removed from model

River channels: units & layout, connections with surface & subsurface land units

Factor	WRSM-Pitman (Sami GW)	SPATSIM-Pitman (Hughes GW)	ACRU4	SWAT2012	MIKE-SHE, semi-distributed, more conceptual	MIKE-SHE, fully-distributed, more physical
Number of channel units by subcatchment (or catchment)	Flexible	1 per subcat	Flexible	1 per subcat*	Flexible	Flexible
	Channels represented with channel modules in a network. No limit to number of channel modules. Multiple connection configurations possible: max 10 input & 10 output routes per channel module.	Subcat channel receives flow from the channels of upstream subcats & flows out to channel of downstream subcat.	Multiple channel units can be included in a network in a subcat. In non-headwater subcats a channel is needed to receive flow from upstream subcats & route to the subcat’s outflow node. Each special riparian HRU & wetland HRU needs its own linked channel unit. No limit to the number of channel units included.	Each subcat has one ‘main channel’ that receives flows from main channels of any upstream subcats & flows out to channel of downstream subcat. Each subcat also has one conceptual ‘tributary’ that routes HRU runoff to main channel. This allows for additional delay & loss to aquifer if relevant.	Channels are represented as a spatially explicit network of reaches between calculation nodes (cross sections) at specified intervals. Reaches with nodes mapped in subcat can exchange water with that subcat. No limits to numbers of river branches or of reach units that can be in a subcat.	Channels are represented as a spatially explicit network of reaches between calculation nodes (cross sections) at specified intervals. Reaches exchange water with grid cells that border them. No limits to numbers of river branches or of reach units in a model. There cannot be two reaches in one grid cell (node spacing compatible with grid)
Surface flow from land units into channel	Yes	Yes	Yes	Yes	Yes	Yes
	Runoff, irrigation, mine modules to a linked channel module	Subcat to its channel	HRU to linked channel in subcat	All HRUs in subcat to tributary then subcat main channel	‘Overland flow’ zones in subcat to reaches in subcat. Series routing: only from most downslope zone	Grid cells bordering a reach to that reach (if flow is over bank height)
Flow from channel onto land surface (overbank flooding)	(to wetland unit only)	(to wetland unit only)	(to special riparian HRUs & wetland HRUs only)	No	Yes	Yes
	A wetland can be included within a channel module.	A wetland can be included at downstream end of subcat.	Special HRUs for riparian areas and wetlands can be included. Each needs its own linked channel unit with an overflow threshold.	A wetland can be included in a subcat, but it is not on the ‘main channel’ or fed by overflow.	Channel flow over capacity is routed onto floodplain surface. This water can infiltrate into flooded grid cells.	Channel reach to bordering grid cells
Interflow into channel	Yes	Yes	Yes	Yes	Yes	Yes*
	Runoff, irrigation, mine modules to a linked channel module	Subcat to its channel	HRU to a linked channel in subcat	All HRUs in subcat to tributary then subcat main channel	Most downslope ‘interflow reservoir’ in subcat to reaches in subcat.	Perched, temporarily saturated layers (‘interflow’) handled in ‘saturated zone’. A channel can receive water from such a layer
Flow from channel into unsaturated zone (transmission loss)	No	No	No	Yes*	No	Yes*
	Channel transmission loss leaves the model	Channel transmission loss only to aquifer	No channel transmission loss	Main channel transmission loss goes to ‘bank storage’ – accessible for ET, not part of subcat aquifers.		Perched, temporarily saturated layers (‘interflow’) handled in ‘saturated zone’. A channel can lose water to such a layer.
Groundwater (GW) flow into channel	Yes	Yes	Yes	Yes	Yes	Yes
	Runoff module to a linked channel module	Subcat downslope aquifer portion to subcat channel	HRU to a linked channel in subcat	Subcat shallow aquifer to channel (not from the deep aquifer)	All ‘baseflow reservoirs’ in subcat to reaches in subcat	Saturated layers of grid cells bordering a reach to that reach
Flow from channel into GW (transmission loss)	No	Yes	No	No*	Yes	Yes
	Channel transmission loss leaves the model	Channel to subcat aquifer (downslope portion)	No channel transmission loss	Main channel transmission loss does not enter subcat aquifer. Tributary can lose to subcat shallow aquifer	Reach in subcat to baseflow reservoir in subcat	Reach to saturated layers of grid cells bordering it
Abstractions: water leave model	Yes	Yes	Yes	Yes	Yes	Yes
Transfers: between channels	Yes	No	No	Yes	Yes* (complex set-up)	Yes* (complex set-up)
External water inputs	Yes	Yes	Yes	Yes	Yes	Yes
Abstraction, transfer, external input locations	Flexible (avoiding circular routing)	Subcat channel outlet  (above reservoir)	Flexible	Subcat main channel outlet (above reservoir)	Flexible	Flexible

Water bodies (dams, reservoirs, lakes): unit types, inflow & outflow connections

Factor	WRSM-Pitman (Sami GW)	SPATSIM-Pitman (Hughes GW)	ACRU4	SWAT2012	MIKE-SHE, All approaches
Water body unit types	1 type*: Reservoir module	2 types*: Reservoir & Dam	1 type: Dam	3 types*: Reservoir, Pond, Depression	2 approaches: explicit bathymetry+wall, storage unit
	Wetland unit also modelled as water body (see below)	Wetland unit also modelled as water body (see below)		Wetland unit also modelled as water body (see below)
Number of units allowed per subcatchment (or catchment)	Flexible	1 Reservoir per subcat & 1 Dam per subcat	Flexible	1 Reservoir per subcat & 1 Pond per subcat & multiple Depressions	Flexible
Flows into water body units	Reservoir module can receive: Rain Flow/transfers from up to 5 other modules routed to it (land areas, channels, & other reservoir modules) - from runoff modules it can receive a set proportion of total total runoff. External source input routed to it A reservoir receiving a set proportion of runoff from a runoff-module can be thought of as internal to the subcat that the runoff module represents. If it receives all the runoff, it’s at the subcat outlet.	A reservoir is on the subcat channel at the outlet, so receives: Rain Local subcat runoff Channel flow from upstream A dam is internal to a subcat & receives: Set proportion of total subcat runoff	A dam can receive: Rain Flow from channels & HRUs routed to it External source input routed to it	A reservoir is on the subcat channel at the outlet, so can receive: Rain Local subcat runoff Channel flow from upstream subcats Transfers from channels & reservoirs in other subcats A pond is internal to a subcat & receives: Rain Set proportion of total subcat runoff A depression/pothole is ponded water ON a specified HRU & receives: Rain Set proportion of runoff from other specified HRUs in same subcat	Waterbodies are handled in the channel hydraulic model (MIKE-Hydro). Waterbody bathymetry cross sections and the dimensions of a dam wall can be input in a reach of the channel. The reach could receive: Rain Runoff from bordering units (cells, zones) Channel flow from upstream reach Transfers from other reaches External point source inputs Alternatively a simple 'storage' volume unit can be added to the end of a channel branch, which receives flow from the branch.
Flows out of water body units	Water can leave a reservoir module as: Evaporation Overflow & controlled release to linked channel module downstream Withdrawal to linked irrigation modules Withdrawal transferred other reservoirs & channels Withdrawal removed from model A reservoir can have a maximum of 5 outflow routes.	Water can leave reservoir as: Evaporation Overflow & controlled release to downstream subcat channel Withdrawal removed from model Water can leave a dam as: Overflow to subcat channel Withdrawal to subcat irrigation sub-area	Water can leave a dam as: Evaporation Overflow, controlled release, & seepage to linked downstream channel or subcat outflow node Withdrawal to linked irrigation HRU Withdrawal removed from model	Water can leave a reservoir as: Evaporation Overflow & controlled release to downstream subcat channel Seepage to subcat aquifer Withdrawal to linked irrigation HRU Withdrawal transferred other reservoirs & channels Withdrawal removed from model Water can leave a pond as: Evaporation Overflow & controlled release to downstream subcat channel Seepage to subcat aquifer Water can leave a depression water body as: Evaporation Overflow & controlled releases to subcat main channel Seepage to local HRU soil (local HRU also has: veg ET & subsurface runoff generation)	Water can leave a reach storage created with bathymetry & a wall structure as: Evaporation Overflow & controlled release to downstream reach Seepage to bordering unit groundwater Withdrawal to linked irrigation area Withdrawal transferred other reservoirs & channels Withdrawal removed from model Water can leave a ‘storage’ volume unit as: Overflow to a downstream branch

Wetlands: units, location, surface & subsurface flow connections

'Wetlands' are challenging to generalise about because a diverse range of hydro-geomorphic settings can give rise to a frequently saturated area. The modelling tools take different approaches to representing wetlands, as described in more detail on a dedicated wetlands page, here. The table below is a brief overview of the connectivity of specific wetland modules or units in the tools. For some wetland types, other approaches besides the specifically named 'wetland' modules may actually be more relevant (e.g. riparian zone representation in various tools, 'riparian HRU' in ACRU4, 'depression/pothole' in SWAT2012)

Component	WRSM-Pitman (Sami GW)	SPATSIM-Pitman (Hughes GW)	ACRU4	SWAT2012	MIKE-SHE, semi-distributed, more conceptual	MIKE-SHE, fully-distributed, more physical
Specific 'wetland' model unit	Yes* (unit within channel module)	Yes	Yes	Yes	No (wetland conditions to be re-created with topography, channel, aquifer, soil, & cover set-up)	No (wetland conditions to be re-created with topography, channel, aquifer, soil, & cover set-up)
Location of wetland	On channel, flexible location in network	On subcat channel, at subcat outlet	HRU + associated channel, flexible location in network	Internal to subcat, not on main channel	Flexible	Flexible
Surface water flow into wetland	Channel flow over threshold (or all)	Channel flow over threshold (or all)	Channel flow over threshold (or all)	Proportion of subcat surface runoff	Surface runoff from upslope zone &/or channel overflow	Surface runoff from upslope cells &/or channel overflow
Wetland outflow to surface water	Outflow to channel module	Outflow to subcat channel outlet	Outflow routed in network (to channel, dam, or subcat outlet node)	Outflow to subcat main channel	Outflow to downslope zones or to channel possible	Outflow to downslope cells  &/or to channel possible
Groundwater flow into wetland	No	No	No	Proportion of subcat subsurface runoff	Capillary rise only	Capillary rise & GW flow possible
Wetland seep to groundwater	No	No	Wetland HRU has own ‘baseflow store’	Seep to subcat aquifer	Can percolate & recharge 'baseflow reservoir'	Can percolate & recharge aquifer layers

Irrigated areas: units, water source options

Component	WRSM-Pitman (Sami GW)	SPATSIM-Pitman (Hughes GW)	ACRU4	SWAT2012	MIKE-SHE, All approaches
Irrigated area model units	Irrigation module	Irrigation sub-area within subcat	Irrigation HRU (special HRU)	HRU with irrigation set-up	Grid cells in 'irrigation demand areas'
Units (explicit crop types) per subcatchment (or catchment)	Unlimited	1 sub-area per subcat	Unlimited	Unlimited	Unlimited
	No limit to number of irrigation modules & all can have different properties. Each can also have multiple crop type sub-area entries: area-weighted averaging of properties is done in the model.	Area can be subdivided into river-fed and dam-fed, but only one irrigation demand set (hence crop type) input.		Irrigation can be set-up for any HRU in the ‘management parameters’ Land cover input & delineation of HRUs occurs separately.	An ‘irrigation demand area’ map input (extents of irrigated areas by scheduling type) determines if a grid cell is irrigated. The separate land cover map input determines the cell’s veg type/crop.
Number of irrigation water sources allowed per unit	1 source per module	2 sources per subcat: 1 source per portion of total irrigated area	1 source per HRU	Multiple sources for an HRU as scheduled	Multiple sources per irrigated area: switch sources with water availability
		The subcat irrigated area can have a % that is river-fed & % dam-fed, but no one place can be fed by both sources.		HRU irrigation schedule can be set-up with different methods & sources at different specified times. For each time entry only one water source can be selected: cannot set to switch between sources based on their water availability.	Multiple potential sources can be specified with an order of preference. If water supply is lacking (or licensed use amount reached) at a higher listed source, irrigation will be drawn from the next source possible in the list.
Irrigation water source type options:
River	Yes	Yes	Yes	Yes	Yes
Dams/reservoirs	Yes	Yes*	Yes	Yes*	Yes*
		Only from ‘dam’ (fed by local runoff); NOT ‘reservoir’ (on channel; upstream inputs)		Only 'reservoir', NOT 'ponds' or 'depressions/potholes'	Only from part of a channel reach (explicit bathymetry & wall), NOT simple 'storage'
Aquifers	No	No	No	Yes	Yes
External source	Yes	Yes	Yes	Yes	Yes
Restrictions on source locations	Flexible (must be different to module receiving return flow)	Only sources within same subcat	Sources in subcat, or dams in upstream subcats	Flexible (not restricted to same subcat)	Flexible (not restricted to same subcat)
‘Return flow’ routing options	To any channel or reservoir module (that is not the source module)	To subcat channel	To a subcat channel, reservoir, or the outflow node	To subcat channel	Downslope cells &/or channel
		Part of subcat-scale runoff handling	Handled same as other HRUs.	Handled same as other HRUs Can add tile drains: also route to channel, changes rate	Handled same as other grid cells Can add tile drains: drain to point in river, changes rate