Hello, thanks for the tutorial. How do you "define your own grain classes" ? This log2 scale is really not convenient and I was not able to find where to customize it. Thanks in advance !
Go to Options-->User Defined Grain Classes (top option) in the sediment editor. There you can change the grain sizes (note-the program only actually uses the mean size) and names. If your material is sand, start with grain class 6 and higher. RAS automatically assumes those first 5 grain glasses are silt and clay regardless of their size. www.hec.usace.army.mil/confluence/rasdocs/rassed1d/1d-sediment-transport-user-s-manual/entering-and-editing-sediment-data/sediment-properties-options/define-grain-classes-and-sediment-properties
Thanks for fantastic series of videos. 1.I have a quesion how you determine this relationship Qs=0.022*Q2.0 ?? 2. If there no any sediment gauge station upstream what will be the most suitable boundary condition? 3. Could we determine a boundary condition using river flow as you did?
If you don't have sediment data, it gets more difficult. I really need to do a video on this, but you basically have 4 options: 1) Collect some concentration data to develop a rating curve 2) Estimate with some sort of regional regression equation or watershed yield equation, preferably calibrated to deposition rates at a reservoir with similar watershed characteristics 3) if you have calibration data, just make the sediment load your calibration parameter (guess and adjust) or 4) you could run what I call a "credibility test" - if your system has historically been in dynamic equilibrium (if for example, you are planning a future modification of a system that is historically in equilibrium) then simulate several decades of the past with a credible transport function, and develop a rating curve that makes the model stable on the reach/decade scale.
There are some cases where the equilibrium boundary condition would be appropriate. Some of those conditions include: 1) it is way upstream of the area of interest, 2) it is not supply limited or armored, 3) it is in a reach that is actually in equilibrium. But I still prefer to convert the equilibrium loads into a rating curve so I have more awareness of and control over the assumptions the model is making.
Sorry Stanford i have a question, i'm trying to do a simulation of an event of the past, a backanalysis, and i know for sure that the river that i study (a gravel bottom channel) have no problem of erosion. The last year a big alluvional event bring in this channel a lot of fine sediment. When i was get to check the situation i can observe that the bottom of the cannel was clearly without accumulation and without erosion. So the real situation is that there isn't erosion and the accumulation of the fine sediment was only out of the bottom of the main channel (in the floodplain and also in the lateral side of the cannel). I put as a max depht 0,1 m and i calibrate the model with the "vol. out cum" of the last section, the only numerical data that i have for the calibration. The problem is that a lot of the deposition is on the bottom of the channel (in the lateral side of the channel and a little bit out too, but not like in the reality) and the erosion sometime arrive to 0,5 m. Have you some extra help to give me? ;) Hope my english is quite ok to understand the situation, your video are awesome!! Hope that you continue this really important job
Bringing a fine pulse into a domain with a gravel bed is actually a pretty tough problem for an "active layer" type model. There can be a couple of problems. We are writing a tech note right now on what we call the "floating gravel" problem - which results in artificial persistent deposition after a fine pulse in a model. But the other problem is that models like RAS that assume "independent grain classes" tend to over predict gravel and cobble transport and under predict sand transport in a mixed bed. Version 5.1 will have a whole suite of hiding functions that will help (we were having a similar problem in one of our models and hiding improved the model a lot). Until then, the Wilcock and Crowe transport function has hiding in it, you could use that if it is appropriate. Otherwise, you may have to artificially increase the size of the gravel on your bed to compensate for model bias. Also, if you limit the thickness of both the active layer (under mixing options) and the inactive layer (initial max depth) fine pulse models tend to perform better.
Could you please explain what do you mean when you said "this system doesnt exist, and should not be use for planning..."? i mean, can i use this information for a real study or do i need to change some steps?
HEC-RAS exists. :) You can use the program with site-specific data (local cross sections and sediment measurements from your river). But the cross sections and sediment data in this video (and provided) are fictional. I made them up completely. They are not from a real river or even roughly based on a real river. So I would not use them for any real study in part or whole. Does that make sense?
@@stanfordgibson thanks a lot sir, I will definitely reach you. But that problem was solved and actually I am new in HEC-Ras, your videos helped me a lot to know a new thing as a beginner.
Check out the two videos I developed on Reservoir modeling. if that is too complicated, you might consider doing it in HMS. The HEC-HMS team is adding new settling basin algorithms. I'm not sure how many of them were in the latest release, but I think some of them are.
The program shows me an error, unrealistic vertical adjustment at * river station * when i run the sediment simulation, can you help me with this issue ? , why this error happen?
Basically, that means your channel filled with sediment very quickly. The most common reason for that is that the time step is too large. The channel doesn't have time to update the hydraulics in response to deposition. But if it is happening at the upstream cross section, it is usually because the sediment load in the boundary condition is too coarse (often because the user used the bed gradation for the load gradation...but the bed gradation is usually much coarser than the load). This could also happen when the sediment load is too high or the transport function isn’t appropriate. I would like to do a trouble shooting series and could probably do a video just on this error message. It is the most common.
@@johan127lespaul Interpolating shouldn't cause this error...but it might make the underlying error more pronounced. So a model that was mostly unstable but used to run, may crash when you interpolate. But interpolation is revealing a problem, not creating one.
however, sometimes interpolation can create an intermediate cross section without a defined channel (particularly when the thalweg switches sides). In a sediment model, you have to use the "chord ties" (those red lines in the interpolation editor) to make sure the thalwegs are connected if they switch sides of the channel - particularly if you have multiple channels.
We will be releasing a beta for 2D sediment in a couple months. We are testing it internally right now. The video is here: ruclips.net/video/tFLy7c-LUvc/видео.html
Max depth (or minimum elevation) is the depth to bedrock or hard pan or some sort of non-erodible boundary. If the system is just alluvial and there is no known hard bottom, choose the maximum depth you think the river could possibly scour under the model conditions and then evaluate the results. If the river hits the "bottom" you set, either make it deeper or reevaluate some of the modeling assumptions.
Stanford Gibson i have a question about it, when i run the sediment simulation, the program show an error, “ error during sediment simulation: unrealistic vertical adjustment at * a river station * , i don’t know what i have to do.
@@johan127lespaul Basically, that means your channel filled with sediment very quickly. The most common reason for that is that the time step is too large. The channel doesn't have time to update the hydraulics in response to deposition. But if it is happening at the upstream cross section, it is usually because the sediment load in the boundary condition is too coarse (often because the user used the bed gradation for the load gradation...but the bed gradation is usually much coarser than the load). This could also happen when the sediment load is too high or the transport function isn’t appropriate. I would like to do a trouble shooting series and could probably do a video just on this error message. It is the most common.
@@stanfordgibson Thank you. So can I define the max depth with the surveying data? For example, firstly I set a number for the entire channel. Then I adjust the max depth base on the comparison of the results to the surveying data cross-section by cross -section? Does this way work? Thanks.
Many thanks Dr. Stanford.
Hello, thanks for the tutorial. How do you "define your own grain classes" ? This log2 scale is really not convenient and I was not able to find where to customize it. Thanks in advance !
Go to Options-->User Defined Grain Classes (top option) in the sediment editor. There you can change the grain sizes (note-the program only actually uses the mean size) and names. If your material is sand, start with grain class 6 and higher. RAS automatically assumes those first 5 grain glasses are silt and clay regardless of their size. www.hec.usace.army.mil/confluence/rasdocs/rassed1d/1d-sediment-transport-user-s-manual/entering-and-editing-sediment-data/sediment-properties-options/define-grain-classes-and-sediment-properties
@@stanfordgibson Thanks a lot for your answer
Thanks for fantastic series of videos. 1.I have a quesion how you determine this relationship Qs=0.022*Q2.0 ??
2. If there no any sediment gauge station upstream what will be the most suitable boundary condition?
3. Could we determine a boundary condition using river flow as you did?
Hello, I describe how to develop sediment rating curves like that in another video: ruclips.net/video/cBr3KkLTW1Y/видео.html
If you don't have sediment data, it gets more difficult. I really need to do a video on this, but you basically have 4 options: 1) Collect some concentration data to develop a rating curve 2) Estimate with some sort of regional regression equation or watershed yield equation, preferably calibrated to deposition rates at a reservoir with similar watershed characteristics 3) if you have calibration data, just make the sediment load your calibration parameter (guess and adjust) or 4) you could run what I call a "credibility test" - if your system has historically been in dynamic equilibrium (if for example, you are planning a future modification of a system that is historically in equilibrium) then simulate several decades of the past with a credible transport function, and develop a rating curve that makes the model stable on the reach/decade scale.
There are some cases where the equilibrium boundary condition would be appropriate. Some of those conditions include: 1) it is way upstream of the area of interest, 2) it is not supply limited or armored, 3) it is in a reach that is actually in equilibrium. But I still prefer to convert the equilibrium loads into a rating curve so I have more awareness of and control over the assumptions the model is making.
Sorry Stanford i have a question, i'm trying to do a simulation of an event of the past, a backanalysis, and i know for sure that the river that i study (a gravel bottom channel) have no problem of erosion. The last year a big alluvional event bring in this channel a lot of fine sediment. When i was get to check the situation i can observe that the bottom of the cannel was clearly without accumulation and without erosion. So the real situation is that there isn't erosion and the accumulation of the fine sediment was only out of the bottom of the main channel (in the floodplain and also in the lateral side of the cannel). I put as a max depht 0,1 m and i calibrate the model with the "vol. out cum" of the last section, the only numerical data that i have for the calibration. The problem is that a lot of the deposition is on the bottom of the channel (in the lateral side of the channel and a little bit out too, but not like in the reality) and the erosion sometime arrive to 0,5 m. Have you some extra help to give me? ;) Hope my english is quite ok to understand the situation, your video are awesome!! Hope that you continue this really important job
Bringing a fine pulse into a domain with a gravel bed is actually a pretty tough problem for an "active layer" type model. There can be a couple of problems. We are writing a tech note right now on what we call the "floating gravel" problem - which results in artificial persistent deposition after a fine pulse in a model. But the other problem is that models like RAS that assume "independent grain classes" tend to over predict gravel and cobble transport and under predict sand transport in a mixed bed. Version 5.1 will have a whole suite of hiding functions that will help (we were having a similar problem in one of our models and hiding improved the model a lot). Until then, the Wilcock and Crowe transport function has hiding in it, you could use that if it is appropriate. Otherwise, you may have to artificially increase the size of the gravel on your bed to compensate for model bias. Also, if you limit the thickness of both the active layer (under mixing options) and the inactive layer (initial max depth) fine pulse models tend to perform better.
can anyone give me sediment transport data for simulation?
Could you please explain what do you mean when you said "this system doesnt exist, and should not be use for planning..."? i mean, can i use this information for a real study or do i need to change some steps?
HEC-RAS exists. :) You can use the program with site-specific data (local cross sections and sediment measurements from your river).
But the cross sections and sediment data in this video (and provided) are fictional. I made them up completely. They are not from a real river or even roughly based on a real river. So I would not use them for any real study in part or whole. Does that make sense?
HDF_ERROR trying to use HDF output file
Subroutine HDF_ATTRI_SCALAR_CH_READ , HOW TO SOLVE THIS
@@stanfordgibson thanks a lot sir, I will definitely reach you. But that problem was solved and actually I am new in HEC-Ras, your videos helped me a lot to know a new thing as a beginner.
Is it possible for you to make a video on sediment modeling in a settling basin?
Check out the two videos I developed on Reservoir modeling. if that is too complicated, you might consider doing it in HMS. The HEC-HMS team is adding new settling basin algorithms. I'm not sure how many of them were in the latest release, but I think some of them are.
@@stanfordgibson Thanks a lot. I will go through it.
The program shows me an error, unrealistic vertical adjustment at * river station * when i run the sediment simulation, can you help me with this issue ? , why this error happen?
Basically, that means your channel filled with sediment very quickly. The most common reason for that is that the time step is too large. The channel doesn't have time to update the hydraulics in response to deposition. But if it is happening at the upstream cross section, it is usually because the sediment load in the boundary condition is too coarse (often because the user used the bed gradation for the load gradation...but the bed gradation is usually much coarser than the load). This could also happen when the sediment load is too high or the transport function isn’t appropriate. I would like to do a trouble shooting series and could probably do a video just on this error message. It is the most common.
Thanks ! Other little question, this error can be generated because of interpolated sections?
@@johan127lespaul Interpolating shouldn't cause this error...but it might make the underlying error more pronounced. So a model that was mostly unstable but used to run, may crash when you interpolate. But interpolation is revealing a problem, not creating one.
however, sometimes interpolation can create an intermediate cross section without a defined channel (particularly when the thalweg switches sides). In a sediment model, you have to use the "chord ties" (those red lines in the interpolation editor) to make sure the thalwegs are connected if they switch sides of the channel - particularly if you have multiple channels.
can anyone give me sediment transport data for simulation?
I am looking for 2D sediments transport videos
We will be releasing a beta for 2D sediment in a couple months. We are testing it internally right now. The video is here: ruclips.net/video/tFLy7c-LUvc/видео.html
can anyone give me sediment transport data for simulation?
I have a question on how to determine the max depth? thanks
Max depth (or minimum elevation) is the depth to bedrock or hard pan or some sort of non-erodible boundary. If the system is just alluvial and there is no known hard bottom, choose the maximum depth you think the river could possibly scour under the model conditions and then evaluate the results. If the river hits the "bottom" you set, either make it deeper or reevaluate some of the modeling assumptions.
Stanford Gibson i have a question about it, when i run the sediment simulation, the program show an error, “ error during sediment simulation: unrealistic vertical adjustment at * a river station * , i don’t know what i have to do.
@@johan127lespaul Basically, that means your channel filled with sediment very quickly. The most common reason for that is that the time step is too large. The channel doesn't have time to update the hydraulics in response to deposition. But if it is happening at the upstream cross section, it is usually because the sediment load in the boundary condition is too coarse (often because the user used the bed gradation for the load gradation...but the bed gradation is usually much coarser than the load). This could also happen when the sediment load is too high or the transport function isn’t appropriate. I would like to do a trouble shooting series and could probably do a video just on this error message. It is the most common.
@@stanfordgibson Thank you. So can I define the max depth with the surveying data? For example, firstly I set a number for the entire channel. Then I adjust the max depth base on the comparison of the results to the surveying data cross-section by cross -section? Does this way work? Thanks.
can anyone give me sediment transport data for simulation?