| Message/Author |
|
|
|
|
From an email enquiry by Nico Batz:
Hi Nico,
Of course I remember you! OK, I'll have a go at answering your questions:
You are exactly right - when I spin up I use the actual data set I may use for the simulations - for example if I have 100 years of discharge then I will run the model - plot the cumulative sediment output - and when it starts to tail off then I would say the catchment/reach has spun up. Its rarely more than 10 years worth of sim as a simple rule. So you are absolutely right when you say:
" wouldn´t it be better to run the spin up run with a normal Q variation over a few years so that all the surfaces will be influenced by erosion?
maybee use the same data of a year for several year but order this dataset from low Q to high and then use it over a few years, so that it gets comparable? or can i do it also over 1 month? i think it depends on how long it takes for the bad to armour, right?"
Just - do as you plan above - but plot the cumulative sedi Q and you'll see a tail off. The plot below is a bad example – as it shows the sediment discharge rising in more than one spot (due to a large flood event 40% of the way through) but you can clearly see that after the first 10% the output stabilises.
Let me know how you get on,
All the best,
Tom
Sent: 08 January 2010 18:41
To: Thomas J Coulthard
Subject: spin-up run CAESAR
hello Tom
i have a question on the spin-up run for the reach mode for a braided river:
i did some preliminary tests, and out of the results it looks like that for braided rivers it might be difficult to create the base grain-file by using a constant Q for a certain spin-up time.
the main characteristic of this rivers (and their existence) is the high variable Q. this influences the distribution of the grain-size within the large active corridor. a constant Q will not influence the entire active corridor and transforming the river in a single treat river and so cleaning the study area just in the river channel.
if after, the base grain file is used as model input suddenly the entire corridor is affected by the high variable Q, flushing away the sediments of large areas that weren´t "cleaned up" by the spin up run before.
so i think that this method works well for the catchment mode (or single thread), because through the rainfall all cells get wet (erosion), but for the reach it doesn´t because influenced by the input Q.
is my thinking right?
do i need to think of an a other method for the spin up run?
wouldn´t it be better to run the spin up run with a normal Q variation over a few years so that all the surfaces will be influenced by erosion?
maybee use the same data of a year for several year but order this dataset from low Q to high and then use it over a few years, so that it gets comparable? or can i do it also over 1 month? i think it depends on how long it takes for the bad to armour, right?
you have any suggestions?
kind regards
Nico |
|
|
 Nico Batz (Nico) posted on Tuesday, February 09, 2010 - 07:42 pm
|
|
|
hello Tom,
i did the spin-up run:
i used a Q spin-up file which starts with max Q of 1290 m³/s, lowering down (exponentially) to a min of 0.34 after 91 days and increase again to 1290 at day 182. So, i have a bimodal hydrograph as in reality. i thought this would be the best option because it takes in consideration the highly variable Q of the river.
the graph shows (hopefully it works) the sediment-Q.
i thought that at day 1643 the spin-up is done.
the method work quite well.
at the moment i´m trying to integrate the lateral erosion.
i just have the problem that at low flow the river is eroding to much and at high flow not enough.
do you know how i can do that?
i´m tryng to keep the TAU velocity threshold as high as possible, so that the lateral erosion gets more sensitive to the flow velocity, and in the mean time lateral erosion low. Problem is that by doing this i got to much incision in the channel.
is there a kind of spin-up for the lateral erosion as well?
you have any suggestion?
Thanks in advance,
Nico |
|
|
|
|
Hi Nico,
SOrry for not replying sooner - I tend not to check the message boards every day so send me an email reminding me to check if you've not had a reply for a day or so..
OK, good to hear you've got the spin up working well
I've moved your question - and the answer to this thread:
http://www.coulthard.org.uk/discus/messages/3/62.html?1265970767
Tom |
|
|
| Back to top |
|
