run.dtm: Run Dynamic TOPMODEL against hydrometric data and a catchment...

In dynatopmodel: Implementation of the Dynamic TOPMODEL Hydrological Model

Description

Run Dynamic TOPMODEL against hydrometric data and a catchment discretisation

Usage

run.dtm(groups, weights, rain, routing, upstream_inputs = NULL, qobs = NULL,
  qt0 = 1e-04, pe = NULL, dt = NULL, ntt = 2, ichan = 1,
  Wsurf = weights, Wover = weights, i.out = ichan[1], dqds = NULL,
  sim.start = NA, sim.end = NA, disp.par = get.disp.par(), ...)

Arguments

groups	Data frame of areal group definitions along with their hydrological parameters (see Metcalfe et al., 2015)
weights	If the discretisation has n groups, this holds the n x n flux distribution (weighting) matrix defining downslope
rain	A time series of rainfall data in m/hr. One column per gauge if multiple gauges used.
routing	data.frame Channel routing table comprises a two-column data.frame or matrix. Its first column should be average flow distance to the outlet in m, the second the proportions of the catchment channel network within each distance category. Can be generated by make.routing.table
upstream_inputs	xts A list of any upstream hydrographs in addition to hillslope runoff feeding into the river network
qobs	Optional time series of observation data
qt0	Initial specific discharge (m/hr)
pe	Time series of potential evapotranspiration, at the same time step as rainfall data
dt	Time step (hours). Defaults to the interval used by the rainfall data
ntt	Number of inner time steps used in subsurface routing algorithm
ichan	Integer index of the "channel" group. Defaults to 1
Wsurf	matrix Surface routing matrix. Defines routing of overland flow downslope between units. By default identical to subsurface routing matrix by default, but can be altered to reflect modified connectivity of certain areas with the hillslope
Wover	matrix Optional surface overflow routing matrix. Defines routing of overland flow from a unit that has run out of surface excess storage capacity. Identical to surface routing matrix by default. Can be altered to reflect an overflow channel for a runoff storage area, for example.
i.out	For multi-channel systems, the index of the outlet reach
dqds	Function to supply a custom flux-storage relationship as the kinematic wave celerity. If not supplied then exponential relationship used.
sim.start	Optional start time for simulation in any format that can be coerced into a POSIXct instance. Defaults to start of rainfall data
sim.end	Optional end time of simulation in any format that can be coerced into a POSIXct instance. Defaults to end of rainfall data
disp.par	List of graphical routing parameters. A set of defaults are retrieved by calling disp.par()
...	Any further arguments will be treated as graphical parameters as documented in get.disp.par

Details

The grouping (HRU) table may be generated by the discretise method and includes each indexed channel as separate group. See Metcalfe et al. (2015) for descriptions of the parameters maintained in this table.

Evapotranspiration input can be generated using the approx.pe.ts method

Value

qsim: time series of specific discharges (m/hr) at the specified time interval. can be converted to absolute discharges by multiplying by catch.area

catch.area: the catchment area in m^2, calculated from the areas in the groups table

data.in: a list comprising the parameters supplied to the call

datetime sim.start Start of simulation

sim.end datetime End time of simulation

fluxes: a list comprising, for each response unit the specific base flows qbf, specific upslope inputs qin, drainage fluxes quz, and any overland flow qof, all in m/hr

storages: a list comprising, for each response unit, root zone and unsaturated storage, total storage deficit and surface storages (all m)

Note

If rain, pe or observation data differ in time period, use aggregate_xts to coerce the relevant series to the desired time interval

Author(s)

Peter Metcalfe

References

Metcalfe, P., Beven, K., & Freer, J. (2015). Dynamic TOPMODEL: a new implementation in R and its sensitivity to time and space steps. Environmental Modelling & Software, 72, 155-172.

See Also

aggregate_xts

discretise

Examples

## Not run: 
require(dynatopmodel)
data(brompton)

# Examine the November 2012 event that flooded the village (see Metcalfe et al., 2017)
sel <- "2012-11-23 12:00::2012-12-01"
# Precalculated discretisation
disc <- brompton$disc
groups <- disc$groups
rain <- brompton$rain[sel]
# to 15 minute intervals
rain <- disaggregate_xts(rain, dt = 15/60)
# Reduce PE, seems a bit on high side and resulted in a weighting factor for the rainfall
pe <- brompton$pe[sel]/2
qobs <- brompton$qobs[sel]

# Here we apply the same parameter values to all groups.
# we could also consider a discontinuity at the depth of subsurface drains (~1m)
# or in areas more remote from the channel that do not contribute fast subsurface
# flow via field drainage
groups <- disc$groups
groups$m <- 0.0044
# Simulate impermeable clay soils
groups$td <-  33
groups$ln_t0 <- 1.15
groups$srz_max <- 0.1
qobs <- brompton$qobs[sel]
qt0 <- as.numeric(qobs[1,])
# initial root zone storage - almost full due to previous event
groups$srz0 <- 0.98
# Quite slow channel flow, which might be expected with the shallow and reedy
# low bedslope reaches with very rough banks comprising the major channel
groups$vchan <- 400
groups$vof <- 50
# Rain is supplied at hourly intervals: convert to 15 minutes
rain <- disaggregate_xts(rain, dt = 15/60)
weights <- disc$weights
# Output goes to a new window
graphics.off()
x11()

# Initial discharge from the observations
qt0 <- as.numeric(qobs[1,])

# Run the model across the November 2012 storm event
# using a 15 minute interval
run <- run.dtm(groups=groups,
               weights=weights,
               rain=rain,
               pe=pe,
               qobs=qobs,
               qt0=qt0,
               routing=brompton$routing,
               graphics.show=TRUE, max.q=2.4)

## End(Not run)

dynatopmodel documentation built on May 1, 2019, 7:32 p.m.