No warrenties, expressed or implied, are made that the computer
programs described in this publication are free from errors or
are consistent with any particular standard of programming language,
or that they will meet a user's requirement for any particular
application. This software cannot be part of any commercial venture
without express authorization by the author.
If you have questions
or remarks about the software on this page (including the PCRaster related
software) please contact: email@example.com
NOTE: the manual is not up to date!
changes in LISEM
openLISEM: Lisem is now open source: go to http://lisem.sf.net for the latest version
lisem 2.5.8-2.6.4 (VJ 100203)
- added multiple interception storage functions for different land use types, after Dejong and Jetten 2007 (default is the original equation)
- added possibility to use a canopy storage map smax.map (mm)
- added user defined stemflow fraction (5.4% default after Bruinzeel, van dijk), stemflow was fixed at 40% since the beginning of LISEM!!!
- added optional "hardsurf.map": hard surfaces that have no infiltration, no interception, no detachment (splash and flow), no surface storage, but they do have runoff, transport and deposition. If set to 1 a cell will behave as like this, if set to 0 the other surface properties will be used. Do not use this for roads, roads are done with the road width map (because a road can be smaller than a grid cell). To use this map use the hard surface option in the main interface.
- Added choice for surface storage: hard threshold with max depression storage (MDS) or non-linear gradual increase of runoff between 0.1*MDS and MDS (default)
- Display of maps while running. Switch off to speed up. Each timestep can be saved as jpeg when the screenshot button is pressed.
- Changed Ksat calibration to a fraction instead of % (1.0 means no calibration is applied). Careful importing old files, calibration may go to 100.
- Added calibration factors now for Manning's n, Channel Ksat and Channel Manning's n.
- Added more clearly conssrvation measures: grass strips, water buffers and sediment traps. Sediment traps are buffers without a water retention, only sediment retention (such as a wicker fence). Do not use for grass strips.
- Added cumulative runoff as map series
- Added soilloss map = sum detachment and deposition
- Deleted runoff water height map series and changed to cumulative water height
- Added possibility for unlimited output of hydrographs (and sedigraph) using a new "outlet.map". Note that the main outlet must have the value '1' and correspond to the ldd main outlet. All other can be just points in consecutive numbers 2, 3 … n. These can be stored as separate files or all in one file
- The ldd map no longer has to be one connected network, more pits are allowed. However only the outlet pit will be showed in the result screen and the water of the other pits is not stored or converted to overland flow. It simply disappears. The water balance is still correct because the outflow of all points is summed together.
- Outflow point hydrographs can be formatted to fit as input into the SOBEK flood model.
- LDD map is now reference map, area.map is no longer used. All maps must have a value where LDD has a value. This is tested when loading the run file.
Bug Fixes and code changes:
- further decoupling of interface and model, model now called with temp run file that is parsed instead of directly by interface, for future open source version
- added 'celltest' function where the cell is found that gives a problem compared to the LDD
(repairs infamous kinematic wave error)
- changed linked list memory declaration in kinematic wave
lisem 2.5.1-2.5.3 (VJ 080620)
- Added snowmelt (EXPERIMENTAL)
- Added baseflow to channelflow (EXPERIMENTAL)
- Added non-erodible surfaces "hardsurf.map" (0 = erodible, > 0 is non-erodible)
- changed sed concentration in all output to be calculated as tot sed/volume of water in the cell not as Qsed/Q
- fixed bug sed concentration to screen
- fixed bug with endless loop when missing map
- fixed small bugs baseflow maps
- improved dir name checking of interface
- no more Smith and Parlange or Morel Seytoux
- Split masbalance correction for water (only gully and wheeltracks) and sediment (all instances)
- deleted screen variable infiltration in kin wave, is always on
- added check for empty name in maplist
- added chanvol in m3 for baseflow estimation
- corrected MC and NUT code
lisem 2.5.0 (VJ 080217)
- Added baseflow to channelflow, mutually exclusive with channel infiltration
- Deleted screen variable infiltration in kin wave
- Added check for empty name in maplist
- Added output map series chanvol in m3 for baseflow estimation
lisem 2.391-2.40 (VJ 170806)
- for safety: inithead should be negative but is read as inithead = -abs(inithead)
- corrected some errors with ginfil in lispgully
- added infil to gullies, similar to channel infil
- bug fixes in Gully kin wav
- bug fixes in GA: green and ampt 2 was not activated and Soildepth1 typing error in stead of SoilDepth1
- bug fix in commandline running, wrong check for lisemtype
- bug fix in runiing with pest: checkpestout, listens to commandline only
lisem 2.36-2.391 (VJ 071005)
- Green and Ampt behaves more clearly now:
- In a normal soil when initial moisture is less than porosity, the soil is seen as infinite and the wetting will continue to move downward and the infiltration rate will eventually decrease to a level of ksat. An exception are cells that are completely saturated, in which case the wetting front equals the soil depth and the infiltration rate does not change . In case of a 2 layer G&A only soildepth1 is used, soildepth2 is seen as infilnite. Overland flow will be Hortonian unless the second layer is very impermeable.
- If "subsoil impermeable" is checked the soil will fill up and after that saturation overland flow will occur.
- If "subsoil drainage" is checked, a gravity driven drainage flux is calculated:
K = Ks*(theta/porosity)^d
where d is a user defined drainage factor > 1. The drainage is applied to the bottom of the profile (soildepth1 or soildepth2) and decreases the moisture content theta. When the infiltration front reaches the lower boundary drainage is assumed to continue as ks and the wetting front depth is decreased.
- Fixed output infiltration maps: infiltration in kinematic wave was not correctly included.
- Fixed infiltration in multiclass and nutrient kinematic wave, this was not correctly done.
- Possibility for running in a PEST automatic calcibration environment added.
- Batch file running improved. Call lisem as:
syntax: lisemwin [runfilename] [lisemtype] <pestout 0/1> <pest time interva>
- lisemwin.ini file is no longer used
- Added units of output erosion in kg/cell; kg/m2; ton/ha
- minimum maning's n decreased to 1e-5.
lisem 2.35 (VJ 050811)
- fixed a rounding of bug in the timeinput (rainfall reading) causing a no data available error for large timeseries.
- fixed a small bug in the writing of the sed. conc. of sub catchment outlets 2 and 3
lisem 2.159 to 2.34 (VJ 050704)
- changed version numbering: major ver 2, minor ver 3, release 0, build = 18.104.22.168
- added channel infiltration = ksat*wet perimeter*dx*dt
- changed: according to new analysis PondAreaFract = 1-exp(-1.875*WH/(RR*10))
- can choose alternative simplified flow detachment: Df=Q*dt(TC-C)*Y instead of Df=w*h*vs*dt(TC-C)*Y
- (added rainwater buffers: but not activated)
- added gully initial width and depth maps and option for equal distribution of erosion
along gully perimeter (instead of Q-W relation)
- removed coh=100 limit in Y
- output map series are now set to min-max of the whole serie
- fixed: bug in infil fluxes into kin wave
- fixed: channels can run with wheeltracks
- fixed: V was not recalculated in channel cells after subtraction of tochannel,
so wrong stream power in channel cells
- FIXED BUG (040224) in area ponding calculation, /10 or *10. /10 gives very smooth
surfaces and area is 100% very quickly, should be *10 (cm to mm)
- fixed: changed Y to ChannelY in cannel calculations
- redone the wheeltrack module
- moved pwheeltrack.cpp to before surface storage because of changes in WH, mass error wheeltracks below now below 1 %
- correction: added kin wave infil check to lisemwin.ini file
internal version: date 030630
- fixed typo: mixed suspended for nh4 and no3 in lisPMCNUTerosdepo
internal version: date 030627
- fixed typo xxxPdep = xxxPDep in all the code
internal version: date 030626
- hanged checking of DirExist in all input dirs
- added max lines for Richedit OutputValues field, -1 = show all, nr > 0 is show nr lines
internal version: date 030414
- FIXED channel bug , factor 1000, introduced in version 2.13
- added nutrient in channels
- added eros and depo totals for nutrients
- redesigned NUT kinematic
- added NUTmassbalsed for calculation of totals
- NOTE: changed calculation of MC sediment and NUTs to flux/flux
WINCVS started in april 2003 !!!
Internal version, limited availability
Integrated the EUROWISE code
with LISEM, now available as LISEM
Bug fix: the memory allocvated
to a rainfall file in a run was not freed. This caused erratic errors
when doing a series of runs at the same time.
Changed the calculation of
the sediment flux in the kinematic wave to an explicit solution of the
sediment mass balance. A description is not yet included on this website.
Added surplus infiltration
to the kinematic wave, so that water that flows inside the kinematic
wave routine can still infiltrate when it flows over a pixel that has
some infiltration capacity left. Read
about it here.
Due to simulations of very
steep terrain the slope has been added to the gridcell length: instead
of dx it is now dx/cos(a) in one direction. This affects many calculations
(see e.g. rainfall
depth) that convert water height to water volmune. The corrected dx is
also used in the kinematic wave.
Fixed a bug that reported a
negative average runoff in the simulation window. This did not effect
the actual simulation, only the output to screen.
Changed the influence of the
settling velocity from a linear function to an exponential decrease function.
Added a minimum water height
of 0.01 mm below which all sediment will be deposited and there is no
Changed the cutoff value for
the deposition: see here.
Changed the upper limit of
the Manning's n map from 0.6 to 10 to enable simulation under forest.
Added different versions of
LISEM to deal with different types of processes (wheeltracks, multiple
sediment classes, nutrients, gullies)
Run file changed.
Changed gradient from tangent
to sine of the terrain slope, in line with literature. This
influences in steep terrain
- the effective rainfall
on the surface,
- the Manning equation for
overland flow velocity,
- the kinematic wave procedure,
- the streampower equation
for sediment transport
Changed surface storage theory
and corresponding algorithms
Fixed a series of small interface
Ksat calibration is added for
all infiltration equations, not only SWATRE.
Snowmelt removed, approach
too simplistic and never worked properly.
Added additional map for the
number of wheeltracks when using "wheeltracks as channels" option:
WHEELNBR.MAP. Thus more than two wheeltracks per gridcell, or rills(!)
can be simulated.
Corrected an error in the sediment
concentrations that were shown on screen as compared to the files (n.b.
only a display bug, the actual calculations were not affected).
Changed the SWATRE infiltration
option: only one option is left which deals with all different surfaces
now, depending on map values
Changed the way grass strips
are dealt with
Changed the run file layout.
LISEM still reads old run files but makes a new runfile which is more
Added a snowlayer throughflow
module, description forthcoming
The Headout option is back:
the matrix potentials in the soil a specified locations can be stored
when using the SWATRE infiltration option. A map HEADOUT.MAP has to be
made with the location id numbers (numbers > 0), similar to the OUTLET.MAP.
Channel infiltration adapted:
it is assumed that the infiltration is saturated and that the Ksat value
can therfore be subtracted from the water in the channel. Also a maximum
storage has to be defined for the channel bottom to simulate a channel
on hard rock. The two maps needed are CHANKSAT.MAP (unit: mm/h) and CHANSTOR.MAP
(mm). The latter is defined as : store = depth*(porosity - initial moisture).
Added "geometric mean" to the options of SWATRE. When checking this box, the average conductivity
K between nodes is calculated as: Kavg = sqrt(Ki *Ki-1)
else the average K is calculated as: Kavg = (Ki +Ki-1)/2
NOTE: the conductivity between the surface and the first node is always
calculated with the geometric mean (Feddes et al, 1978).
Added "calibrate Ksat" to the options of SWATRE. The value in the box can be used to adjust Ksat:
if the value is 100 the Ksat is not changed, 90 means 0.9*Ksat, 110 means
1.1*Ksat. The adjustment is done for all layers and all mapunits. The
unsaturated conductivity is not changed.
version prior to 2.33
Channel infiltration is still
Channel infiltration still
doesn't work properly!
Diagonal connections option
should not be used, it doesn't work properly.
Fixed: "floating point
error" on SWATRE precision input menu .
Fixed: the output name of the
discharge map series, mapname was always n0010.000 when regular output
interval was chosen .
Fixed: the rainfall input file
was not changed if more then one run was done.
The older DOS versions 3.x
and 4.x have a serious bug in the calculation of the overland flow discharge
between gridcells! DO NOT USE THESE VERSIONS.