RRM Model: concepts and processes
The main elements, concepts and processes represented in the model are:
- pre-processing RRM is a spatially lumped model, but considers different land use classes
(defined with the basin parameters) and elevation bands (defined in a special
to adjust some of its parameters as weighted averages.
- Precipitation: the raw precipitation time series data are adjusted using
two optional parameters and the elevation distribution: a plain multiplier, directly applied,
and a precipitation correction factor, expressed as a % increase (or decrease) of precipitation with
every 100m altitude layer; the correction is pre-processed as a weighted average over the elevation
bands and their relative size, then applied together with the basic multiplier.
- Temperature: raw temperature (time series) input can be adjusted with shift (in degree Centigrade)
and an adiabatic lapse rate, that defines temperature decrease with altitude. From that, a corrected temperature
is being computed, again as an elevation weighted average.
- Interception storage: this varies with land use/land cover; the available total interception
storage capacity (in mm) is computed as a weighted average over the land use classes.
- Evapotranspiration: is again land use class specific; a weighted average degree day
factor is derived from the land use distribution, and the land use specific parameters.
- Infiltration: is a function primarily of land use; again, an area weighted average
is being built from the class specific parameters.
Interception storage, which is a function of land cover;
excess water reaches the soil (surface);
from the interception storage, evaporation (function of temperature) is
Soil surface; depending on infiltration capacity, (depending on land
cover and soil moisture), precipitation is split into surface runoff
(Hortonian flow) and infiltration.
Depending on the air temperature (profile), which is calculated from the
basins elevation distribution, precipitation can be both in the form of
rain or snow. In the latter case, a snow pack is simulated,
from which both evaporation and melting (the latter greatly enhanced by
precipitation events above freezing) are estimated using a simple
The rootzone; this compartment represents the soil moisture in
the rootzone, which is replenished up to field capacity;
from it, evapotranspiration (a function of temperature, land cover, and
soil moisture) returns water to the atmosphere;
water in excess of field capacity reaches the next storage level, the
Virtual drainage storage. From here water is split into an
interflow component and percolation into the first groundwater layer.
Assuming a maximum speed of percolation for water transfer to the shallow groundwater layer,
any excess water is routed through the unsaturated soil system to the nearest
channel as interflow.
The (two layer) groundwater system supplies, in a non-linear
function of storage, the baseflow contribution of the total runoff.
A fraction of the (upper, shallow) groundwater can percolate into a second,
deeper layer that can also contribute to baseflow.
The percolation is specified a mm/m/day of the storage level of the shallow aquifer.
From the second, lower deep aquifer, another contribution to the baseflow
(usually rather slow and with considerable delay)
can be modeled, using a response lag (interpreted as half time) specified in days:
360 days would mean that half the groundwater reservoir would drain into the
base flow contribution in this period with an exponential pattern.
From both shallow and deep groundwater, groundwater extractions (time
series defined for specific wells or well fields) can be considered.
The channel system using Muskingum routing to move the water
through the basin to the outlet.
The initial conditions to be specified include:
together, they they define basin water content (water in the channel is being ignored)
which is used in the overall (annual) water/mass budget.
- snow pack (in mm water equivalent);
- basin state (interception storage, soil moisture);
- groundwater content (shallow and deep reservoirs).
- Initial snow pack is specified in mm water equivalent;
its distribution is assumed to be exponential with 50% in the top layer,
and again half (of the rest) in each lower layer.
- Basin state: to simplify the specifications and ensure consistency,
this is done in symbolic terms. The initial state of the basin can be:
- very wet (interception storage full, 100% soil moisture)
- wet (interception storage half full, 90% soil moisture)
- normal (interception storage empty, 80% soil moisture)
- dry (interception storage empty, 60% soil moisture)
- very dry (interception storage empty, 30% soil moisture)
- Groundwater: shallow and deep reservoirs,
specified in mm of water content (water equivalent, not considering aquifer porosity).