Playback speed
×
Share post
Share post at current time
0:00
/
0:00
Transcript

Internship Opportunity on CFD ,MCDM in Flood Inundation Mapping

Call for Internship: An internship program where we want to use CFD and MCDM techniques to predict the impact of flooding in a watershed by mapping the area that can be inundated by flood.

The Baipatra VSC newsletter discusses an upcoming internship program focusing on the application of computational fluid dynamics (CFD) and multi-criteria decision-making (MCDM) in flood prediction studies. The program aims to use CFD and MCDM techniques to predict the impact of flooding in a watershed by mapping the area that can be inundated by flood. CFD is often used in hydrology for detailed flow analysis, flood international mapping, and flood inundation mapping. It can also be used to access the impact of proposed flood control structures on water flow and inundation levels, allowing for optimized design and placements. CFD can also accurately simulate urban flooding, helping to identify areas with poor drainage and potential flooding hotspots.

The internship program aims to integrate the advantage of multi-criteria decision-making techniques with CFD, as CFD lacks the ability to segregate variables based on their role in flood prediction and cannot generate maps that provide information about the extent of potential flood damage. By integrating CFD with other advanced techniques like MCDM, GIS, or nature-inspired optimization techniques, the application of CFD will be more optimal and effective in flood prediction studies.

CFD (Computer-Aided Data) is a powerful tool for accurately forecasting flood dynamics, inundation zone mapping, and comprehensive flood risk assessment. It is also used for training rescue staff and enhancing accuracy and computational efficiency. 3D CFD simulations are now common for flood risk analysis, and researchers are using it to simulate real-world flooding scenarios. CFD can also access the convergence of numerical solutions using techniques like the grid convergence index. However, there is a lacuna in its use, as it requires a lot of data and is often based on numerical schemes. CFDs are theoretically validated but not tested in practical situations.

One concern is the accuracy of CFD models developed with only CFD, as they cannot measure the temperature inside the surface area of an iron plate. Integration with other technologies like MCDM or GIS is rare, and only some limited research is available. Digital twin technology is being used for flood risk management.

The discussion revolves around the limitations of Computer-Aided Flood Forecasting (CFD) in predicting watersheds. CFDs are primarily used for visualization purposes, but they cannot be verified due to inaccessibility. To address this, research can focus on improving data availability and accessibility through investments in data infrastructure, capacity building, and promoting open data initiatives. This can attract large-scale investments.

One major limitation is that CFDs require data, which can lead to erroneous results if there is a data scarcity. Additionally, most CFD models are single-dimensional, which requires a significant computational infrastructure, which is not available to developing countries or poor researchers.

Another option is integrating different modeling approaches like MCDM, GIS, and remote sensing to create a comprehensive flood risk management framework. Digital twins, which are 3D replicas of watersheds, can be developed using 3D technology, but this requires significant investment and virtual goggles.

In conclusion, the discussion highlights the need for research in improving data availability and accessibility, integrating different modeling approaches, and exploring emerging technologies like digital twins to enhance decision-making processes.

The speaker is interested in a research project that aims to identify the exact parameter responsible for Flood Level Attribute (FLAT) and use advanced MCDM techniques to predict flood proneness in a watershed. They believe that most studies only consider water level and elevation, but this would be a complex system that requires large computational infrastructure. The speaker plans to develop a single parameter that embeds the weighted contribution of all input parameters related to FLAT, allowing them to predict inundation areas and flood proneness using advanced MCDM techniques.

The speaker is considering using the analytical hierarchy process and Promethi and a recent technology called MERC, M-E-R-C. They invite interested candidates to apply for an internship at their lab in Tripura, North East India, which is available throughout the year and lasts two months. The outcome of the internship will be a certificate and a chance to publish a paper in a reputed journal or conference. The speaker will consider all applications uniformly without bias and encourages interested candidates to send their CVs.

(00:00:01):

Hello and welcome to yet another podcast of Vaipatra VSC newsletter.

(00:00:12):

Today I want to discuss about the internship program which is going to be launched next year.

(00:00:26):

The internship is about

(00:00:31):

the application of CFD,

(00:00:33):

computational fluid dynamics,

(00:00:36):

and multi-criteria decision-making,

(00:00:39):

MCDM,

(00:00:41):

in flood prediction studies.

(00:00:45):

Specially,

(00:00:47):

we want to use CFD and MCDM techniques to predict the impact of flooding in a

(00:00:55):

watershed by mapping the area

(00:00:59):

that can be inundated by flood.

(00:01:05):

So,

(00:01:05):

as all of us know,

(00:01:07):

computational fluid dynamics is used in flood prediction to simulate the complex

(00:01:12):

flow dynamics of water in rivers and floodplains,

(00:01:17):

allowing researchers and engineers to accurately predict flood inundation areas,

(00:01:25):

water depths,

(00:01:26):

and flow velocities,

(00:01:28):

which can be crucial for earlier warning systems,

(00:01:32):

mitigation strategies,

(00:01:35):

and infrastructure design in flood-prone regions.

(00:01:42):

The title of the internship is proposed as,

(00:01:45):

Application of Computational Freed Dynamics and Multicradiation Making Techniques

(00:01:51):

in Flood Inundation Mapping.

(00:01:56):

So generally, in the field of hydrology, CFD is not used frequently.

(00:02:03):

But when it is used, it serves the following applications.

(00:02:10):

The example application can be detailed flow analysis.

(00:02:15):

CFD models can capture intricate details of water flow, including turbulent effects.

(00:02:22):

water surface profiles,

(00:02:24):

and interactions with obstacles like bridge and buildings,

(00:02:29):

providing a more comprehensive understanding of flood dynamics than simpler models.

(00:02:36):

Second example can be flood international mapping, which we want to do as well.

(00:02:45):

Though we have a difference, we want to do something different, as will be discussed later on.

Share

(00:02:53):

Flood inundation mapping.

(00:02:58):

This flood inundation mapping is by simulating water levels across a geographical area,

(00:03:05):

CFD can generate detailed inundation maps,

(00:03:10):

identify areas at high risk of flooding,

(00:03:13):

and help prioritize mitigation efforts.

(00:03:18):

Okay.

(00:03:19):

So that is why we will try to do some flood inundation mapping for a local watershed.

(00:03:27):

It is a mesoscale watershed.

(00:03:30):

And okay,

(00:03:32):

I will give the details later on,

(00:03:33):

but let me see what other objectives can be achieved by using CFD.

(00:03:42):

CFD can be used to access the impact of proposed flood control structures

(00:03:48):

like levees flood walls or drainage systems on water flow and inundation levels

(00:03:55):

those impacts can also be analyzed allowing for optimized design and placements so

(00:04:01):

accordingly you can place those hydraulic structures such that minimum scouring or

(00:04:07):

minimum erosion or minimum corrosion can be ensured and also extreme flood

(00:04:15):

scenarios can be analyzed or simulate

(00:04:18):

with the help of CFD like sudden dam breaks or rapid snow melt to understand

(00:04:24):

potential worst case scenarios and plan appropriate emergency response strategies.

(00:04:32):

Now,

(00:04:33):

in addition to that,

(00:04:34):

CFD is often used for urban flood modeling by incorporating complex urban features

(00:04:41):

like drainage networks and buildings

(00:04:44):

CFD can accurately simulate urban flooding,

(00:04:47):

helping to identify areas with poor drainage and potential flooding hotspot.

(00:04:56):

What is the significance of applying CFD and MCDM in flooding and national mapping?

(00:05:00):

Now, I have said that we have a catch.

(00:05:03):

We have something different.

(00:05:05):

We have something you can refer as unique selling point in this objective of internship.

(00:05:13):

The USB of our internship is we want to apply CFD.

(00:05:19):

That is very okay.

(00:05:21):

Others have done it, though rare, but they have done it.

(00:05:24):

They had applied CFD in inundation mapping of flood, but we want to integrate the advantage

(00:05:32):

of multi-criteria decision-making techniques.

(00:05:36):

Why?

(00:05:38):

Because CFD lacks the ability to segregate variables based on their role in flood

(00:05:46):

prediction and cannot generate maps that provide information about the extent of

(00:05:58):

potential flood damage.

(00:06:01):

Right?

(00:06:02):

And on the other hand,

(00:06:03):

when MCDM and GIS are integrated with CFD,

(00:06:07):

these limitations can be addressed effectively.

(00:06:11):

Now we will say, why GIS?

(00:06:12):

We are just talking about MCDM.

(00:06:15):

So,

(00:06:16):

why I refer to GIS is,

(00:06:22):

I want to say if CFD is applied by integrating with other advanced techniques like MCDM,

Share

(00:06:31):

GIS,

(00:06:32):

or nature-inspired optimization technique,

(00:06:36):

then the achievement of that application will be more optimal,

(00:06:42):

more beneficial compared to when we are using only CFD.

(00:06:46):

That is my point.

(00:06:48):

Okay.

(00:06:49):

So,

(00:06:50):

you know,

(00:06:51):

just CFD is used to accurately forecast flood dynamics,

(00:06:57):

inundation zone mapping,

(00:06:59):

comprehensive flood risk assessment,

(00:07:02):

They are even useful for training rescue staff and are being developed to enhance

(00:07:08):

accuracy and computational efficiency,

(00:07:12):

right?

(00:07:12):

Because CFD ensures accuracy, okay?

(00:07:17):

3D CFD simulations are now common for the flood risk analysis that are communicated

(00:07:25):

to various public and private authorities.

(00:07:30):

Then safety is utilized nowadays for dam break modeling.

(00:07:34):

Right?

(00:07:35):

So potential dam failures can be identified or dam breaches during storms can be

(00:07:41):

identified before that storm has taken place.

(00:07:47):

Now the many researchers have also used 3D hydrodynamic models to simulate real world flooding scenarios.

(00:07:56):

Helping to choose suitable models for dam break analysis.

(00:08:00):

So the entire dam break can be visualized in a 3D environment.

(00:08:05):

How?

(00:08:06):

By the help of the CFD.

(00:08:09):

Right?

(00:08:10):

And CFD can also access the convergence of numerical solutions using techniques

(00:08:16):

like the grid convergence index.

(00:08:19):

Right?

(00:08:19):

Various numerical solutions can be done with the help of CFD.

(00:08:25):

Because CFD itself is a

(00:08:29):

generation of solution or approximation of solution in a spatial domain.

(00:08:37):

It is the only technique which can predict or which try to solve partial differential equation spatially.

(00:08:52):

Due to this property of CFD, we can use this CFD for flood irradiation mapping, 3D visualization, etc.

(00:09:01):

But, you know, there is a lacuna.

(00:09:06):

What is the lacuna?

(00:09:08):

Lacuna is that, you know, the CFD requires a lot of data.

(00:09:19):

Okay?

(00:09:22):

CFD requires a lot of data and another thing is that CFD is often they are often

(00:09:38):

based on some numerical schemes and mainly they are theoretically validated but

(00:09:47):

practically they are not validated.

(00:09:50):

They are not

(00:09:52):

experimented in practical situations.

(00:09:55):

So theoretically, CFD is very good.

(00:09:59):

When you apply it in your computer, CFD can do anything.

(00:10:04):

You give the temperature of four sides of iron plate.

(00:10:10):

CFD will predict what is the temperature in the surface area of the iron plate.

(00:10:17):

But whether the temperature predicted by the CFD

(00:10:22):

truly exist in the surface, that is a point of concern.

(00:10:29):

Because generally you cannot measure the temperature of a plate within the surface.

(00:10:42):

By surface I am not meaning that at the surface area, I am meaning what is inside the surface.

(00:10:50):

Right?

(00:10:50):

Not on the surface.

(00:10:52):

I am saying inside the surface.

(00:10:55):

Okay?

(00:10:56):

So you cannot measure that by any instruments.

(00:11:04):

And there is no empirical model available for predicting that.

(00:11:11):

So as a result, we have to depend upon this type of CFD technology to approximate or to estimate

(00:11:20):

to measure what is the temperature inside the surface area of the iron plate.

(00:11:32):

So the accuracy cannot be verified for the models developed with only CFD.

(00:11:41):

So another thing is that it is very rare when CFD is integrated with other technologies like MCDM or GIS.

(00:11:52):

And only some limited research we can identify like some of the latest digital twin

(00:12:00):

technology is used for flood risk management.

(00:12:05):

But there it is just a visual representation of the watershed in a 3D environment.

(00:12:11):

You cannot verify the whatever flow the CFD is approximating for those areas where

(00:12:18):

you cannot go and measure the flow.

(00:12:22):

that value you cannot verify with anything because they are inaccessible.

(00:12:34):

So what to do?

(00:12:36):

Right?

(00:12:38):

So that is why as per the research gaps in the previous studies, you can do your research in this field

(00:12:53):

on improving data availability and accessibility through investments in data infrastructure,

(00:13:00):

capacity building and promoting open data initiatives.

(00:13:04):

I remove this promoting open data initiatives.

(00:13:08):

You can either trade or you can develop your data infrastructure.

(00:13:14):

Both of them will attract large scale investments.

(00:13:19):

Because generally, CAB used

(00:13:22):

CFD are implied to approximate something which is unmeasurable.

(00:13:28):

If it is measurable, then we could have used it.

(00:13:32):

Then you see developing advanced models that can accurately represent complex

(00:13:37):

psychological processes and urban environments even with limited data.

(00:13:42):

You know, one of the major lacuna was CFD requires data.

(00:13:51):

So if there is a data scarcity, results from CFD would be erroneous.

(00:13:58):

Okay, and also most of the model of the CFD is single-dimensional.

(00:14:06):

There are very few models where 3D is used,

(00:14:09):

but 3D is mostly used for visualization purpose,

(00:14:13):

not for analysis purpose.

(00:14:16):

Okay, so, and if you want to use or develop some model X, Y, Z,

(00:14:22):

you have to calculate it or approximate it in all three directions.

(00:14:30):

And that will require a lot of computational infrastructure,

(00:14:34):

which is not available to the developing countries or poor researchers.

(00:14:41):

Another thing that you can consider is integrating different modeling approach

(00:14:49):

like MCDM, GIS, remote sensing to create a comprehensive and robust flood risk management framework.

(00:14:59):

This is the most easiest job that you can do.

(00:15:06):

Why?

(00:15:08):

You can also, as I say, the data predicted by CFD cannot be validated.

(00:15:13):

So you can develop something for that also.

(00:15:16):

and also you can explore potential of emerging technologies like digital twins to

(00:15:22):

enhance part modeling and decision making approach processes.

(00:15:26):

But as you know, what is digital twins?

(00:15:30):

Digital twins is that the exact replica of the watershed can be generated in a 3D system, right?

(00:15:40):

Suppose those, you know the goggles,

(00:15:44):

3D goggles that are used for virtually going to space or virtually going to some forest.

(00:15:53):

Right?

(00:15:54):

So if you visit some tourist spot also, they are doing that.

(00:15:59):

Virtually they are taking you to many different places.

(00:16:03):

So this is what is known as digital twins.

(00:16:06):

So the technology will be embedded in the paper.

(00:16:09):

By scanning the technology, you will go to the

(00:16:14):

exactly to that watershed where the exact replica of that watershed is made with

(00:16:19):

the help of 3D technology and for 3D technology there they use CFD

(00:16:24):

But for that you need the virtual goggles, virtual 3D goggles.

(00:16:30):

And also a lot of investment is required to convert a 2D or 1D simple paper-based

(00:16:36):

picture to a corresponding 3D interactive imaginary twins of that picture.

(00:16:47):

Okay, so it is very, very, very expensive.

(00:16:51):

So I am not interested in fifth.

(00:16:53):

I am not interested in fourth.

(00:16:55):

What I am interested in third,

(00:16:57):

because first of all,

(00:16:58):

it doesn't attract money or doesn't attract much investments.

(00:17:04):

Secondly, I am not considering GIS or considering remote sensing care.

(00:17:09):

I am considering only MCDM.

(00:17:12):

Why I am considering?

(00:17:13):

I am considering because I want to

(00:17:17):

identify the exact parameter which is the only or mostly responsible for FLAT.

(00:17:27):

If I can identify that parameter or if I can use all the related parameters as a

(00:17:34):

single parameter where the weighted contribution of all the input parameters

(00:17:40):

related to FLAT is embedded,

(00:17:45):

then

(00:17:46):

I can use the CFD techniques to predict how it is spreading or it is behaving within a surface.

(00:18:03):

Okay, so my main aim is that what I am going to do is, most of the studies you will see

(00:18:09):

They will use water level and they will try to see what are the elevation of that area.

(00:18:15):

If your water level is more than your elevation, then it is under flood.

(00:18:19):

If not, it is not under flood.

(00:18:22):

But they do not consider the impact of infiltration.

(00:18:27):

They do not consider the impact of evaporation.

(00:18:30):

They do not consider the impact of depression that are already available in the watershed.

(00:18:37):

Because they will reduce the water level that is increased due to the excess water

(00:18:42):

coming into the watershed.

(00:18:45):

Right?

(00:18:46):

But yes,

(00:18:47):

if I want to include all of them,

(00:18:49):

then it will be a very complex system,

(00:18:52):

attracting huge or requiring huge computational infrastructure.

(00:18:57):

To avoid that, I want to develop a single parameter.

(00:19:04):

where the weighted contribution of all the reliable parameters will be embedded

(00:19:11):

right and this this parameter will be used to predict the flood proneness of our

(00:19:24):

area and with the help of that i can easily predict the inundation area or

(00:19:31):

inundated area due to that flood

(00:19:35):

And all these weights will be estimated with the help of advanced MCDM techniques.

(00:19:44):

I am considering to apply our very popular analytical hierarchy process.

(00:19:53):

And another very rarely used technique in this type of cases is

(00:20:05):

Promethi and a very recent technology known as MERC, M-E-R-C.

(00:20:17):

So if you want to know or if you want to experience that this research objective to

(00:20:25):

come into reality,

(00:20:28):

then you are heartily welcome to my lab for doing this internship under me

(00:20:35):

And my lab is located in NIT Agartala.

(00:20:41):

It is a place in Tripura, North East India.

(00:20:46):

But if you face difficulty in coming to Tripura from your place,

(00:20:51):

then we can arrange a virtual internship also.

(00:20:57):

And for that,

(00:20:58):

kindly send your interest and CV,

(00:21:03):

short CV,

(00:21:05):

to this mail,

(00:21:05):

contact at the redenergyinstyle.website.

(00:21:14):

Or you can chat with me as well, as you wish.

(00:21:19):

What I can say is this internship is available throughout the year.

Share

(00:21:23):

But the duration is two months.

(00:21:25):

Within two months, this work can easily be done.

(00:21:28):

And as the outcome of internship, I will give you a certificate.

(00:21:32):

And also, I will give you a chance to publish a paper in a reputed journal or a conference.

(00:21:41):

So it will depend on the performance of the candidate joining the internship.

(00:21:47):

So all of you are welcome.

(00:21:50):

Send your CVs.

(00:21:53):

And accordingly, I will select the most suitable one.

(00:21:59):

And as this is available throughout the year,

(00:22:03):

you can apply anytime,

(00:22:05):

provided this internship is already taken by someone.

(00:22:09):

I will, of course, consider all the applications uniformly without any bias.

(00:22:17):

Thank you.

(00:22:17):

Thanks for your patient hearing.

(00:22:20):

I will sign off now.

(00:22:25):

Good night and I hope some of you at least will email me with their interest to do the internship.

(00:22:37):

Thank you.

Share Baipatra VSC

Baipatra VSC
Baipatra VSC Podcast
Very Short Term Courses on Hydroinformatics Engineering. Once subscribed expect to receive free for all tutorials on various soft computation techniques and their applications on water resource development.