You are browsing as a guest. Sign up (or log in) to start making projects!

FLOWS

  • 10 Devlogs
  • 35 Total hours

FLOWS: Flood Likelihood from pre-compiled Watershed Structure.
Make flood-risk assessment with low-end hardware! FLOWS finds bowl-like depressions and analyses them. Why is that good: We now know were water accumulates, and for a fixed amount of rainfall max water height in theory should stay about the same.
This means we can run a one time computation that analyses different scenarios of rainfall and soil humidity and store the values so we can check them later anytime!

Ship #2 Pending review

Okay, so this is FLOWS part 2! This ship basically includes everything that my previous shit didn’t have, like a github page and an executable file with a DEMO on releases. I also took the chance to fix some bugs that came along the way, and implement changes that were needed while compilation evolved:

  • Excessive RAM usage when generating images fixed (so far, we got to 160.000 km² without crashing - huge improvements)
  • New starting point variable: Stop the code at any time and continue were you left: just type the basin id from where you want to start.
  • Code now can be stopped without loosing previous data
  • A DEMO function is now available on the executable file
  • 5 devlogs
  • 6h
Try project → See source code →
Open comments for this post

27m 8s logged

Alright, so new changes to FLOWS! Basically what was happening was: Some HUGE basins showed up, and the program was exceeding my humble 8gb of RAM when trying to generate a diagnostic image (that also happened on cuts where there were thousands of basins). BUT after some tinkering, I discovered that I was doing some very heavy processes on image generation that were not necessary. So I adapted my code, and now a whole terrain cut of 160k square kilometers (about 39 million acres) can be processed without crashing!

0
0
3
Open comments for this post

56m 4s logged

Alright, so new changes! Today I tried to make a DEMO for my project! Since the user is originally intended to provided topological files (that for demonstration purposes may not be available), I have also packaged inside the .sh file a DEMO version, that has its own very small dataset, so everyone can see the project in action even without the needed files.

0
0
19
Open comments for this post

37m 7s logged

Another day and guess what: another feature! I added a new function to the code that takes in a “starting point” for basin compilation. What this does is, given a starting point integer provided by the user (call it S), it skips all basins with an ID lower than S. This is good ’cuz we can stop the code with control + C and later keep on where we left by choosing our last compiled basin as our starting point! I probably should have done this a long time ago instead of right know when I actually needed it, but hey: Better late than never am i right?

0
0
17
Open comments for this post

2h 39m 18s logged

YEEEEAAAAH! I decided to try and create an executable for my script, and boy was hard. The problem was that major libraries like gdal and whitebox required a bunch of other system libraries to work. This means I had to try to get them into the executable, and let me tell ya: This took a lot of time. But finally, it works (i hope)! I will be posting the executable (linux only of course) on github later!

0
0
105
Open comments for this post

1h 14m 4s logged

Update on compilation: Well, a lot of stuff happened. I can’t really keep my computer running non stop all day and night, so I was resorting to suspending the laptop so runtime would keep frozen in RAM. But today, after a week of running the program and almost halfway left, I woke up to find the computer out of battery (probably it kept running in suspend mode and just died). That was quite devastating, but the only thing I could do was to update my code with checkpoints so that if that happens again, I won’t need to run everything again like I’m about to do now. I took that chance to also fix some typos in my README and my github pages, so I guess I got something good out of this!

0
0
34
Ship #1 Pending review

This is FLOWS. It is a program that takes in terrain data and water infiltration data and finds bowl-like depressions (or simply “basins”). Why is that useful: for each basin, water-height doesn’t change (for the same rainfall). This means we can pre-calculate various rainfall scenarios and store their results, so we can have a simple and quick way of estimating flood risk.

Compilation was also made focusing on lower-end hardware, were we avoid feeding the entire data into the RAM, and instead partition the dataset in smart ways so we can make it all fit within our constrains!

  • 5 devlogs
  • 29h
Try project → See source code →
Open comments for this post

4h 51m 3s logged

Just discovered this nice tool called WhiteBoxTools, and was also able to find its user manual! Now I won’t need to process terrain from scratch, and instead get I can all the data I’m looking for without reinventing the wheel.

0
0
7

Delete project?

Are you sure you want to permanently delete this project? This action cannot be undone.

All devlogs, followers, and associated data will be removed.

Followers

Loading…