Comment by nemo1618 | Hacker News Reader

Comment by nemo1618 | Hacker News Reader

Hi all! Some background on the project: I want to build a house someday, and the first step is deciding where to build. I knew an interactive map like this would be useful, but after fiddling with leaflet.js for a while and googling for climate data, I felt a bit lost. So I decided to try something new: I hired a contractor on Upwork who had experience with map visualizations. They built an MVP within a few weeks, and from there I was able to tweak and optimize the code to my liking. I think this is a great strategy for side projects that you're too busy for. :)

It's still mostly a toy at this point, but I'd like to add more knobs and data so that I can narrow down the best place to live (and hopefully help other people do the same). If you want to contribute, the code is open source: https://github.com/lukechampine/goldilocks

The site weighs about 1.6 MB (gzipped -- the full climate data is ~35 MB uncompressed!) and is hosted on a Linode, which seems to be holding up pretty well.

Also, a big thank you to those in the comments who shared links to similar projects! The "pleasant places" map is super close to what I want. Ideally this project would look something like that, but with the ability to select your own preferences for what constitutes a "perfect" day.

This is awesome. You'll probably want to at least cross-reference this with the dew point, as temperature is only part of the story when it comes to being a comfortable human. Humidity has a large effect on how we perceive a given temperature.

As someone who is considering my next place in the world to live for a 5-10 year period, it would be amazing if this website could support world data, not just the US. If needed, I'm ready to help fund the expansion to support more parts of the world :)

Check out Climate Finder: https://nomadlist.com/climate-finder

Ummm this costs like $100 to use...

Great job! Very responsive.

I thought about making something like this too, years ago. I wanted to converge data from tornados, fault lines, flooding, and hurricaine probabilities. But then I got sucked into EVERY parameter I might want to consider, like air traffic (vortac lines), distance to power grid, UPS coverage, internet coverage, cellular coverage... and eventually the amount of data crushed my brain and the project never even made it to a NextJS boilerplate. :) So if you want to add more data... hint hint hint. :)

How did you "bend" those hexagons? What's going on there as I go from coast to coast? Is that to map them to a sphere?

I honestly have no idea how the heck grid works haha. Maybe someone else can chime in.

The MVP site took >1 second to rerender when you dragged the slider. I got it down to a few dozen ms, which makes it a lot more fun to use. Pretty proud of that :)

The responsive, low-latency controls are definitely a pleasure!

Thank you for not making it a typical CRUD app that queries an overloaded 35MB database for a couple round trips every time you want to move a slider....

Not the dev, but these are probably hexagonal bins that were computed in spherical/conic space (e.g the Albers Equal Area or Lambert Conformal projections [0], which happen to also be the projections that NOAA distributes a lot of their data in) and projected to Web Mercator (which is the web maps projection). Usually when you do that you get bendy things.

[0]: https://psl.noaa.gov/data/narr/format.html

I think it would be very hard to compare the risk from living near a fault line to living in a tornado zone. Tornadoes come nearly every year and it's pretty easy to quantify your risk of being in the path of one (and it's generally easy to minimize the risk -- listen for Tornado warnings and move to a shelter (build one if you have to)).

But the risk from earthquakes is much harder to quantify since they are so rare, hard to predict except in broad generalities, and the damage suffered is also hard to predict. You can try to reduce your risk by following good earthquake practices, but it's hard to quantify the risk reduction.

In my area, the earthquake risk is: “We estimate the probability of having a magnitude 9 earthquake in the next 50 years is about 10 to 14 percent”, but no one can say exactly where the earthquake will strike or what areas will suffer the most damage.

Please add a Celsius / Fahrenheit toggle :)

done!

You can add a Kelvin one as a nerdy easter egg. ;)

You have a great idea, hiring out someone on Upwork for an MVP that you could then run with or not. Wondering, what was the cost?

About $800 IIRC. I didn't shop around at all though.

Leveraging the global brain. Love it.

I can imagine a day where we can do this in our heads and we become a loose super organism.

But I digress, massively!

Thank you for this!!!

I went through scraping weather averages and graphing them in elk, then finding https://weatherspark.com/map and now this. Best option yet.

Well...did it work for you? Did you find some spot where you're going to build a house?

Ha, no, not yet. Looking more closely at NC though. I want to have a nice mountain view, so maybe I need to incorporate topological data too somehow.

NC is a very temperate and beautiful place. Good prices and jobs too. If you live in the middle, both the ocean and mountains are 3 hours away.

Temperate is in the eye of the beholder :-) Pretty hot and humid in summer for my taste. That said, Asheville in particular often gets pretty highly ranked among small cities to live in and it's probably more comfortable in the summer than many other locations in the state.

Asheville is also one of most expensive cities to rent/buy in NC

In general, there's probably a fairly high correlation between places that are more pleasant to live in (across multiple dimensions) and housing prices. Jobs, of course, being another major factor.

This sounds super cool! I'm currently in grad school in an entirely different subject, but I have experience in (spacial) data vis and web scraping. I've used it for some side projects probably few people would find interesting, but that help me in my daily lives. Would upwork potentially enable me to do similar work for clients? Does anyone have any recommendations?

this is cool - i've spent way too much time just sort of synthesizing this sort of data in my brain with no hard evidence.

One knob I'd love to have is the ability to tune the # of days ranges - "above 200" is a pretty large bucket depending on how you tune the other numbers. It looks this is changeable in the code, but maybe it's tied tightly to the available data?

When you hover or click on a hexagon, it would be helpful to have the color it is highlighted in the legend.

Really great! In addition to humidity which others have mentioned, average sunny days per year might be another good metro. From a temperature standpoint it looks like Seattle is ideal for me, but I’m fairly certain I couldn’t handle that few sunny days.

Neat project! Would be nice if it integrated data like wildfires, mudslides, floods, storms/winds, crime, mosquito populations, to be a truly useful aid in home site locating.

Really cool. Does this have a climate change model? Without one this is useful for maybe a 5 to 10yr window.

hugged to death it seems.

Reply on news.ycombinator.com