BMI-type measure for a place's "goodness of weather" | Hacker News Reader

BMI-type measure for a place's "goodness of weather"

7 pointsprofsummergig17 comments

If you're familiar with BMI (body-mass index), read-on:

I am trying to develop a single score, similar to BMI, for a place's "goodness of weather".

Yes, I am well-aware that BMI is flawed. But its simplicity is useful.

The data available to me are:

- MAX TEMP (°F)

- MIN TEMP (°F)

- AVG TEMP (°F)

- PRECIP (IN)

- SNOW (IN)

- Standard Deviation of each of the above.

Data are available for tens of thousands of weather stations in the USA, and are available in the following bin-sizes: annual, monthly, daily, hourly.

I think I want to use monthly data.

BMI is calculated as: weight in kilograms (kg) divided by height in meters, squared (m2).

In your opinion, what should the formula for "goodness of weather" look like?

(Of course, people will disagree on what constitutes good weather. Some like San Diego weather, some like Denver weather. So, what would your formula look like?)

View on HN

You just need to decide your own personal hardiness zone, and compare it to the map.

Just like plants thrive in certain zones, you need to consider what you need in order to thrive.

https://planthardiness.ars.usda.gov/

Those values are not independent variables, so some of them are superfluous (like snow strongly correlates with min and avg temp).

But "good weather" is different for a farmer, a surfer, a skier or a city dweller.

Good points.

Humidity is missing from your list. As is wind chill. Really, you could almost make a score based on just those missing 2, because I can live with a great variety in all the others if I know those two.

Good point. I'm guessing these 2 must be measured by the weather stations. I'll check the variables available.

Update: dew point and wind chill are collected by the weather stations. So yes, I will now aim to incorporate them.

Humidity is massively important to comfort. At least as important as anything else / temperature.

Relevant metric: https://en.wikipedia.org/wiki/Wet-bulb_globe_temperature

Not sure I have an opinion on the actual formula, but I do think there are some other data that would be useful here.

Precipitation frequency should also be considered. For example, Seattle is considered a "rainy" city, because it rains so often, not because it rains so much.

Likelihood of major weather events is also a factor for a lot of people. I'm not a fan or tornadoes and hurricanes, so I'd prefer to live far away from places where those things have a reasonable chance of occurring.

I don't think you can accurately identify "good" weather for even a modest group of random people, because people want different things, for different reasons.

More likely would be identifying locations with similar weather/climate - or even partially similar;

- X has weather very similar to Y; - A has weather similar to B, but gets colder in winter.

Imagine trying to define a "goodness" of food. It's impossible. But identifying similar dishes is much easier and much more useful.

Fantastic points. As I got deeper into the project, I started to realize the futility of it. And have been leaning towards finding "doppelganger" places, sister-cities (regions). The way you framed it ("A has weather similar to B, but gets colder in winter"), has clarified the project for me.

Dump precipitation, add hours (or minutes rather) of sun .

London gets very little precipitation compared to Miami but gets 1500h of sun which is about 1300h of sun less than Miami which gets huge showers and then 2hrs later you are ready to hit the beach again.

Also some of my favorites climates:

Sydney, Cape Town, Aruba, Mexico City, Cabo San Lucas, Cotè Azur, Canary Islands

You lucky person, you've seen a lot of cool places. You're right about hours / minutes of sun. I'll check if the weather stations collect that data.

not everyone of them, some just have good weather stats that entice me to go one day

people will disagree on what constitutes good weather

There is no bad weather, only the wrong clothes.

https://weatherspark.com/ might help you with arriving at some sort of climate scoring system.

if I were creating a BMI for weather, another data point I would add would be max UV index. And ofc, have my 'ideal temperature' (I'll say 70) and then a max and min preferred temperature. And then maximum precip and snow inches. And then the formula for me would be how far each of those preferred numbers are from the places actual numbers (e.g. how far below my preferred min, above my max etc.)

Give me a breakdown of how many 9am 30 minute commutes I'll get reasonably rained on (20%+ precipitation) divided by normalized sun flower hours during the non-winter 3 seasons, please.

Reply on news.ycombinator.com