| Message/Author |
|
|
|
|
| Notes on version 5.8 |
|
|
|
|
20/7/07
Version 5.8 released. Main changes include:
1. Lateral erosion. Two new boxes to control the lateral erosion. First one is length of smoothing. This controls how many times a smoothing filter is passed along the edges of the channel - and therefore determines how smooth (or not) the curvature is for the lateral erosion. This is at this stage experimental, and the larger the number the smoother, and less tight meander bends develop. 100 seems to be a good number at the moment, less than 30 seems to result in tight/notchy bends. There is a good paper in WRR by Crosato that looks at this issue of smoothing the radius of curvature in a more standard Howard type meander bend model. Also, as it smooths the edge curvature term, it reduces it, so will reduce rates of lateral erosion. You may have to increase the lateral coefficient correspondingly.
Second box is the downstream ofset of the edge calculation. If this is set to 0, then there is no downstream migration of meander bends. The number corresponds to how many grid cells (in the downstream direction) the edge curvature calculation is offset by. 1-5 seem to be a reasonable number at the moment.
2. Slope processes tab.
We have moved the creep and landslip angle boxes to here, and introduced a kind of creep/wash term as well. This is based on slope, slope length (root of drainage area) and an exponent or two. If it runs for a long time it results in slightly concave slopes. It might be possible to calibrate/parameterise this term from RUSLE if needed.
The main change here is the addition of the SIBERIA boxes. SIBERIA is a landscape evolution model developed in the early 90's by Garry Willgoose (Newcastle University, Australia). There is a substantial body of literature surrounding it, so I shall not describe its operation here. However, this part of the model is not in any shape or form tested (apart from a brief test) nor compared to any results from SIberia itself. If you have any comments on this part please let me know.
If you wish to use it, tick the siberia box and on the files tab uncheck BOTH catchment and reach mode tabs. This will cause CAESAR to ONLY operate in SIBERIA mode. It can operate with the usual CAESAR flow model etc.. if you check the catchment tab as well - what the results of this would mean I dont know!
3. Google earth support. Run the model, press the export to KML option and CAESAR will save a .png image of what is displayed on the main CAESAR screen, and will also save a .kml file (both files saved in the main directory where the CAESAR files are saved). Double click on the kml and it will open your image in GOogle earth. This was coded up by Jorge Ramirez - thanks Jorge, excellent work.
There is a slight error in getting the transform right for the far RH edge - something to be worked on and its more noticable for large DEM's but for now its fine. There is also an animate to google earth set of boxes that will create a FOLDER containing a kml file and a sequence of .png's dumped at the prescribed time step.
The png;s can also be used for output into documents etc.. or even in an animation program if you need. |
|
|
|
|
OK, - I forgot the changes that were in 5.7 - which of course are also in here. There is also the save as velocity vectors option, that allows you to save a point file with the resultant velocity and an angle. See: http://www.coulthard.org.uk/discus/messages/6/54.html?1184249426
for more details
There are also some more boxes added for the tracer functions - information on these is to be updated. |
|
|
|
|
Notes on the google earth part:
If you have rotated your DEM so that water exits on the RH side, then it will not be 'unrotated' when you export it to Google Earth - sorry. I guess we have to work on being able to select which edge to let water flow off of. etc..
Tom |
|
|
|
|
New method for by passing calibrating the hydrological model if you have discharge data.
On the file tab, there is now a check box marked "if checked jmean is read direct from rainfall file"
If you take your discharge record at the outlet - and divide it by the number of cells within the catchment - then replace the rainfall data file with these values, then it will re-calculate j_mean (the output from the hydrological model) as a proportion of the total discharge. Therefore if you have the discharge data for your catchment (in reality) then you can get CAESAR to accuraely model the flow output. |
|
|
|
|
Charts/plots
On 5.7 and 5.8 there is now the option to plot the discharge and sediment discharge as the model is running. This only plots the amount output at the time the graph is sampled, so it only gives instantaneous flow and sedi readings (not hourly totals for sed) but does give a good indication of what is happening |
|
|
|
|
Hello,
I've been checking the code for CAESAR and seem to have found a bug that will probabaly effect most versions from v5 onwards.
Its quite important, but you need to check whether or not it makes a difference to your simulations - it may well not.
To explain the error: for a while, I have set the code up to use mean bed slope to calculate the slope to determine flow depths. The function (mean_bed_slope(x,y)) has not been returning the correct values - it was returning the max, or the last +ve slope it checked.
This same slope function was used to determine Tau for erosion when using bedlsope.
As such it may effect all simulations, but more so those model runs using bedslope to calculate Tau.
I'm really sorry about this - and I hope that it has not effected your simulations... I thought it best to let you know ASAP.
I would advise checking them against the latest version of the code which I shall post on the website later on today (5.8g) Good luck, Tom |
|
|
| Back to top |
|
