Hacker News new | past | comments | ask | show | jobs | submit login
Show HN: The best time to visit any city
395 points by ignostic on Aug 22, 2017 | hide | past | favorite | 118 comments
I wanted to build a tool to help people decide when and where to travel. As I started building, I realized that "when" and "where" need separate treatment to be most useful. The map tool handles "where" best:


Clicking through each week would be frustrating for those who know where they want to travel but not when. For these people I built "best time to travel" pages using the same data.


My hope is this site will help travelers plan.

This data is taken from the National and Atmospheric Administration's global summaries of the day (NOOA's GSoD). I used an SQL database to crunch the numbers into monthly and weekly averages by station. For the "best time" pages I used and calculated several more variables. I then imported the data into Tableau and added the filters you see on the map. I also used data from the State Department regarding travel advisories.

Would love your thoughts!

The whole buildout was a solo project, but I owe Ryan Whitacker a big "thank you" for his guidance. He built a similar tool on his site (https://decisiondata.org/the-best-time-to-visit-anywhere/) in April, and was generous to offer me guidance for expanding upon his idea.

Known issues:

* I am aware that the map is bad on mobile, so my next step is to improve the mobile experience.

This is super interesting idea and something I'd reference a lot. The big question I have is what are your factors for deciding the best time to visit? I'd argue setting up the context is really important because weather is a major factor but it isn't the only one. There are times of year that have cultural significance as well as annual events, etc. I like the safety advisory aspect and the population of travelers. It would be interesting to know what types of groups travel there and when. So for instance, I want to go to Hawaii but not there are going to be tons of kids and I'd like to do it cheaply and I don't care about the weather.

Lastly, I don't recommend bucketing NZ under Australia :)

Yes! There's a few cities I visit very often but I ONLY visit in the "off season." Everything is much cheaper, things don't sell out, I don't get bothered by stupid events I have no interest in, parking is free or reduced price, I don't have to sit in traffic, and crowds are non-existent. I'd rather avoid crowds than pretty much any other factor other than maybe price.

http://offpeak.io the project my team has been working on aims to solve this exact problem. We help travelers avoid peak season, and peak dates that are often associated with large business conferences.

It seems to be missing all but the largest cities. My guess is that you are manually adding data which would mean higher quality at the cost of general usefulness.

I used an Instagram heatmap to find places to visit/look at during a holiday, I would guess if one adds a "time heatmap" one can see which weeks/months are quieter and which are busier.

The search is really frustrating. Everything I type is not there and then it shows me things that are very far away from what I typed. It may be better to show a list organized by region or show the closest location you do cover to the one that was typed in.

That's great! May I ask where you're getting your data?

I had a great trip with a group of friends to Iceland in mid-winter, which was a time that other people were reluctant to go. It definitely made certain travel activities harder or impossible, but we got to experience lots of great things and many prices were lower. We also saw the spectacular Reykjavík new year's fireworks.

But I'd also appreciate this site's advice is handy because weather in a distant place is so unfamiliar and can have such a big impact on a trip.

Another idea: maybe look through each category in the Köppen classification and figure out what weather factors in such a climate would most affect the quality of someone's visit. For example,


Then somehow highlight those particular factors for each place. (I don't really know how best to organize this information.)

I thought the same thing. My biggest factor is cost of travel and whether its in season or out of season. Then I guess the site would turn into a travelocity or expedia.

> what are your factors for deciding the best time to visit?

I used almost exclusively weather data. My scoring process is linked to near the date ranges I provide, so hopefully that is clear.

I would love to make better use of the number of people visiting. I had a really hard time getting additional travel data for what times are busiest, but I did my best with Google data. I'd even be willing to pay for tourist data by week or month, but I don't believe it exists on the scale I'd need it. If I find it, though, I'll be sure to iterate!

> Lastly, I don't recommend bucketing NZ under Australia :)

Good call. I changed the heading to "Australia & Oceania"

I doubt there's a single dataset containing tourist occupation across the world, but you can probably find it city by city, although often only in the original language.

For example, there's a report on the rate of occupation of hotels per fortnight in Barcelona, but it's a PDF and it's not even in Spanish (Castellano), but in Catalan, the regional language: https://www.diba.cat/documents/74348/115651101/Informe+Flaix...

I wrote earlier that I used a website witb an Instagram heatmap to see places around me in a tourist destination that people took pics, if that website used time as another dimension (not just geo coordinates), it would be a rough approximation of the popularity of destinations per week/month.

Nice. It's true NZ is small but it's worth standing on its own.

Weather is important but it's not the only factor or the most important one for everyone. If there was "One tool to rule them all" for travel I think this could be really powerful. :)

"As of 6/22/2017 there are travel warnings for the United Kingdom; exercise a high degree of caution."

I thought this was an error at first, but the .ca travel advice is indeed to be extra careful visiting the UK.

Spain, meanwhile: "Exercise normal security precautions".


I guess it's only as good as the data sources :)

Terrorist attacks are not earthquakes. Barring further evidence, there's no reason to believe Spain is any more dangerous today than it was two weeks ago. If anything, it's less, since there's one fewer group planning an attack.

Then again, I recently moved to Brussels and I'm planning an impromptu trip to Barcelona thanks to the falling flight prices, so maybe I'm just suicidal :)

You made my point better than me. Ppl go to Germany for Oktoberfest, too Vermont for foliage, to Utah for skiing - additionally, these could be reasons to not go to those places at those times.

Do you think NZ should be bucketed under Asia instead?


Kiwiland is further from mainland Asia than pretty much anywhere else is, excluding the Americas. I'd go for Oceania, and throw Australia in there as well.

Possibly Oceania.

It's a nice idea, for sure. I'm not sure I'd use it much since other considerations apart from weather also exist.

A few things I've noticed:

• The search could use some awareness about other names of locations. For example, München cannot be found, Munich can. Tübingen doesn't seem to exist at all (maybe too small).

• For people outside the US (yes, they exist) it'd also be nice to have a site-wide switch to metric. This then won't require you to have two charts of everything either (except snow coverage which doesn't seem to exist in metric).

• The legends for the charts look a lot like buttons, which can be a bit confusing at first. Maybe it's better to integrate the legend into the charts, e.g. maybe like http://hypftier.de/temp/2017-08-23_090140.png – would also save a bit of space; whitespace currently looks a bit haphazardly applied in general.

• The animation of the charts seems a bit pointless, considering that they're all below the fold anyway.

omg the metric one is so important. it should even be default 8-o

Depends on where most of your audience comes from.

This is really neat! Only thoughts: it assumes one's definition of 'best' is 'best weather' and that in turn means it matches with what your algorithm decides is most pleasurable (which I think it does a pretty good job at). This is probably ok for most people but e.g. I like to travel to see nature and rare species of plants and whatnot, and that completely changes what 'best' is for me as it makes weather not one of the top things to consider.

Thank you! I am very glad others find the algorithm finding the "best times" to be accurate as well. This was definitely the hardest problem to solve as some places have two seasons, some have one, and I'm trying to find the best time on a scale relative to that place.

I'm aware not everyone will be satisfied the "pleasant weather" parameters I used, but I don't think there's another way to give a range quickly. Perhaps I could add some options on those pages. I'm not sure it would be worth the extra clutter and work to catch edge cases, though, as I'll never satisfy everyone's travel preference. Right now those looking for other weather preferences can use the map, but I understand clicking through several weeks is less than ideal. Hopefully the graphs on the page offer a bit more guidance for you.

One question: are you aware of any dataset that would help locate rare species? I can't guarantee I'll use it, but I love having lots of data handy.

I'd prefer not to have an app to find rare species. They're rare for a reason, and that reason is because there are few places the can live, and those places are likely difficult for people to get to. Please don't make it easier for people to get there.

GBIF, https://www.gbif.org/, aggregates biodiversity occurrence data (observations, where specimens held in museums are from, etc).

Everything is available through an open API, although we don't index threat status. Perhaps we should.

Example: Ceratotherium simum, Southern White Rhinoceros [1]. (The threat status shown on the page is taken from an IUCN API.) The distribution data is available either as individual occurrence records [2], or as a summarized map vector tile [3], or a PNG map tile [4]. You would probably want to add a year range filter for recent observations.

You can also download e.g. everything with coordinates and crunch through the data yourself.

You will often find the coordinates have a low accuracy for endangered species, to prevent misuse. I won't comment on whether making such an app is a good idea or not.

(I am a software developer at GBIF.)

[1] https://www.gbif.org/species/2440880

[2] https://api.gbif.org/v1/occurrence/search?has_coordinate=tru...

[3] https://api.gbif.org/v2/map/occurrence/density/0/0/0.mvt?tax...

[4] https://api.gbif.org/v2/map/occurrence/density/0/0/0@1x.png?...

Wow, thanks for mentioning this. Didn't know it but it looks pretty awesome especially becuase of the combination of all datasets.

I'm not sure it would be worth the extra clutter

Not likely. I mean, for like 99% the current parameters are ok, stuff is easy to use, why bother changing it :]

are you aware of any dataset that would help locate rare species

Not directly but there's sites like observation.org, not sure if they have APIs. Be warned though: it's a lot of information because there are so many species so making crawling through that handy and useful would require quite the interface.

I was surprised with the cities shown on the map -- nothing from Britain or Ireland, only Odense from Scandinavia, yet five places in Moldova and loads more in Ukraine.

Looking at Copenhagen [1], the Celsius graph is maxing out at 10°C -- perhaps it would be neater to show a single graph, with a Fahrenheit scale on the left and Celsius on the right. Or just detect that my browser locale is not en_US, and show Celsius...

Minor thing: metres per second (m/s) is a fairly common wind measurement unit. And it should be km/h, not KPH.


This is great - I think the search functionality could be improved a bit (a lot of results following the one I was looking for that didn't seem related, but the right one did show up first).

I wonder if there would be a benefit of a "community" element as well, as in allowing comments on the pages, to give locals the opportunity to chime in with their advice.

Thanks! I agree on the search functionality. Right now this is just a WordPress search, but it would probably be wise to use Google's location API to help find the right place or the closest place to it. I have no experience here, but I can work on adding it.

Will have to think about the community option as well. I'm not sure what the best implementation would be, but it's a good idea. Thanks for the positive response and constructive feedback!

Love the idea, but yes, the search needs some work. For example, the search engine doesn't get "St. Louis, MO" right at all.

Iterate and carry on!

I'd be curious to know why so many search functions have trouble with St. Louis. I was recently trying to book a flight there, and often had to use the airport code to have it show up!

A tiny bit of feedback...

> June – August is slow/unreported season for tourism in Edinburgh, so lodging and other accommodations may cost as much as usual.

This is... odd. In August, Edinburgh has the legendary (Fringe) Festival, a month in which the cities population quadruples, making it easily the most intense month of the year for tourism.

Maybe the dataset requires some manual tweaking?

Yeah, you've uncovered one where the average month in Google was not reported. Generally that means few people searching. In a few cases, sadly, that may be because the hotels are all full and there's no point to searching.

As I've said elsewhere, I think the solution to this is better data on travel volume. I have a lead thanks to this thread on some better data, which is very exciting for me and exactly what I was hoping for in HN.

Yeah, this is exactly my point. Knowing what kinds of groups are traveling to a certain area and when would be largely influential.

On the 'Weather in Sydney' page: "The warmest time of year is generally mid-January where highs are regularly around 61.9°F (16.6°C) with temperatures rarely dropping below 50.3°F (10.2°C) at night"

I think something is off here. That sounds like our winter weather! I would guess our average summer temp would be closer to 30deg C

Likewise Perth has the same error. "The warmest time of year is generally mid-February where highs are regularly around 63.7°F (17.6°C)". The text in Overall Perth Travel Experience by Season is also wrong.

I like the concept though.

"The best times to visit Australia for ideal weather are January 1st to December 30th"

Another interesting tidbit.

Little-known fact: December 31 is "Stabbing Day" in Australia.

"The cull", aka death from alcohol poisoning on NYE.

Weather is just one small piece of people's thought process when planning a vacation. But this tool is clearly all about weather... Perhaps finding a new way to describe it other than "the best time to visit" would avoid people coming down on the tool because they want to talk about more than the weather.

Yeah, as someone from Southern California, you don't really need to think about weather when you come visit... maybe a bit of june gloom, but even that isn't bad.

I love your weather summaries. Your formula nails it with this one: https://championtraveler.com/dates/best-time-to-visit-san-di...

Most weather summaries seem to miss that early/mid-September is significantly hotter than the traditionally hot months of July and August for most locations.

https://championtraveler.com/dates/best-time-to-visit-louisv... > When can you find snow in Louisville? Weather stations report a bit of annual snow likely to be deepest around March, especially close to early March. Powder (new snow) is most often falls around November 12th.

Seeing powder forecasts for Louisville, KY cracked me up.

Thanks! Most places actually give far more specific date ranges, but San Diego is so temperate just listing the months made more sense.

"Powder" I guess is not the word I should be using in places like this, but the dates are usually accurate in places that get powder at all :)

However, there does seem to be a mistake in the precipitation section:

"If dry weather is what you’re after, the months with the lowest chance of precipitation in San Diego are December, February, and then January."

Good eye. @ignostic, it's reporting the highest percent chance of precipitation by month, not lowest, in the textual summary of both cities.

"The lowest chance of rain and snow" blurb is correct.

This seems really off:


It's says the daily highs in the summer are low 60s and that the driest months are in the middle of the winter.

Also looking at it for Madrid:


It's completely wrong. Highs of 22c in August? Try 40c!

If you can fix the data then this will be super useful!

Also, maybe it would be a good idea to let people add their own tips (that people can then vote up/down), because it may be useful to know, for example, that in August in Madrid most bars, restaurants and shops are closed as everyone flees to the coast.

I agree. Check Atlanta for example. The note says, "The warmest time of year is generally late June where highs are regularly around 71.6°F (22°C)" [1]. Meanwhile the historic average highs for August range from 86-89°F [2]. The average low in August [3] is almost as high as what this site says is the warmest regular high.

[1] - https://championtraveler.com/dates/best-time-to-visit-atlant...

[2] - https://www.accuweather.com/en/us/atlanta-ga/30303/august-we...

[3] - https://en.wikipedia.org/wiki/Atlanta#Climate

> It's says the daily highs in the summer are low 60s

Temperature numbers show the average for the week. E.g. the "high" on the dot is not the highest average day, but the average high. I double checked and these look correct. Check out the "heat index" adjusted numbers below for perceived temperatures, as Portland feels significantly warmer due to humidity in the summer.

> the driest months are in the middle of the winter.

I'm confused here. If you're looking at Portland, OR the driest months are shown as being in the summer, so I'm not sure what you're referring to. Did you maybe look at another Portland?

Ah, weekly average makes more sense.

As to the second part:

"If dry weather is what you’re after, the months with the lowest chance of precipitation in Portland are November, December, and then February."

Best time of year is subjective!! All Lonely Planet guide books solve this problems by explaining what each place is like at each time.

Eg, for Yosemite, it describes how during summer it is rediculosy busy & hot, so you might prefer to go during the shoulder months, but then you risk having some of the park closed for snow. So, "best weather is subjective".

Likewise, there are many locations, such as Thailand, where the best time to travel is winter. It's too hot in the summer!!

> there are many locations, such as Thailand, where the best time to travel is winter.

"Winter". Haha. Terminology aside, you're right of course. I was surprised that this actually showed the 'correct' time for most pleasant weather in Ayutthaya (Thailand) - Dec to Feb, the cold/dry season. (People here call it winter, but Thailand doesn't have four seasons so I hate using that term for it)

Weirdly, it suggested almost exactly the same time of year to visit Adelaide, South Australia, which is frankly ridiculous. It'll be hot and dry, and everything will be brown. Seriously.

If you must visit the 1970s. Sorry, I mean Adelaide.. go in Autumn (March - May).

Yes, we travel a lot and what is warm for you might be cold for me etc.

I have solved this problem analyzing more than 20 billion weather data points and created a nice search around it, so you can find the perfect spots that fit your needs depending on temperature, humidity etc. right now it's an internal page for my own use and one of my side projects that I never officially released but use frequently myself.

There seems to be demand so I will put it online in the next days. It will take the subjective out of the best time to go.

Hey, this is really useful! I'll bookmark it and probably use it.

A few bugs/glitches, though:

- The Celsius temperature graph for Barcelona ( https://championtraveler.com/dates/best-time-to-visit-barcel... ) doesn't show temperatures above 25ºC, so every temperature above that gets cropped to 25. The scale should probably be adapted to the data.

- "The busiest month for tourism in Barcelona, Spain is May, followed by March and March." March and March?

- Maybe this is a problem with your dataset and not the app, but just in case, check the snow graph/data for A Coruña: https://championtraveler.com/dates/best-time-to-visit-a-coru... - 108 cm of snow in April? I can guarantee the real average is close to zero :)

- Also in little known tourist destinations (e.g. A Coruña, from the last link) tourism in all seasons is reported as "slow or unreported". Which is true, of course (in the best month in A Coruña you will see much fewer tourists than in a really bad month in Barcelona). But maybe relative data (tourism related to the average in that city) could make sense?

Keep up the good work!

Very helpful, thank you. I will look into these.

Anyone here might be interested in https://weatherspark.com/

Looks cool, but only Fahrenheit.

There's a switch in the top right

Really neat application of the data. I'm the Director of Marketing at a mid-sized destination marketing organization (DMO) represented on your map (Aurora, Illinois -- we're actually a terrific place to visit in winter!). Curious to learn more about the variables used to deploy the "best time" pages and get a sense for where you'd like to take this further.

Recommendation: Travel for snow related trips (snowboarding/skiing) would largely depend on average snowpack on a mountain at that time of year, which is directly related to the amount of snowfall in that location as well as the temperature (does it get cold enough for artificial snow making). Build the right tool for planning snow trips and it should be easy to monetize.

This is a really interesting idea - the kind of tool that you only realize that you've been waiting for when someone shows you it! As someone who likes to travel quite a bit, this will replace a lot of Google searches for me potentially. Your weather summaries will also be great for SEO and are very well optimised.

One bit of feedback I'd give is it'd be worth populating with some "temperate" defaults (i.e. normal average temperature, normal humidity etc.). At the moment it seems like it might take a bit of configuring to get to the information you'd want, when it'd seem like you could take an educated guess.

UI/UX wise I would also make the "date" slider a bit more prominent and maybe have it simply limited to monthly averages (doesn't seem to matter too much if we're talking 2nd or 3rd week of March) for instance.

Otherwise, really love it, and excited to see the ways people are using Tableau for this kind of thing :)

The 'when to travel' could be made dependend on where you live. When I lived in Norway I thought 25C is quite warm. Now that I live in south Germany I think 25C is more intermediate. Also in Spring, temperatures feel a lot higher (since I got used to cold winter weather) then by the end of summer (when I got used to hot weather).

So the best time to visit Sydney is almost the entire year, except for the last 2 weeks of January?

> February 5th to January 14th


Love the idea.

Many of the temperature charts have an uptick at the extreme right (end of December), and then (if you imagine them wrapping around to January) have a non-continuous drop at the start of January. I think there's something slightly funky there.

I saw at least one celsius chart where the data overflowed the maximum (20C) and the plot mashed against the top of the frame.

You could avoid having to double render the same data (and wasting space) by putting celsius on the right y axis of the farenheit chart.

Once you fix the bugs, pay a UI designer to give all of the pages a refresh. Work some SEO, and see if you can find a way to give search engine bots all of the various city pages (which I assume you dynamically generate). Throw in some hotel/airline-ticket site affiliate links and you should get a nice stream of income from this.

Pretty cool job.

It would be nice if the user had a few sliders to toggle (heat, humidity, rain, and crowds), rated say 0 to 10 (with 5 meaning don't care), to get around having to select an ideal temperature for everyone. Some people want sun, others want snow, and others don't care about either.

Nice website, as I had always been looking for such travel website like this.

On the travel-weather-map site, I searched for "San Francisco" and the first result was Argentina, and second was Costa Rica. I hope that the result is based on popularity and not based on alphabetical order.

The search needs some work :) Sadly it is not popularity-based at all.

> Tourism graph is based on Google searches for services used by tourists relative to the rest of the year.

That's a pretty clever way to go about it. I hope/suppose that the sampling bias isn't correlated with time-of-year somehow.

It seems pretty off, though. For example, the page for Iceland says that winter is the peak time, but it's actually pretty dead in the winter with the peak in the summer. Seems like people start planning their summer trips around December (plane tickets for Christmas?). Other places seem shifted up a month or two - the page for Paris has the peak season around May-June, but it's actually more like June-August.

OP, if you want to improve this, could you get some actual tourist data and experiment a bit with which other metrics (various Google searches - something like taxi services might be more accurate to estimate how many people are actually traveling in the country at the time, hotel prices, plane ticket prices, etc) best predict the tourist numbers?

It's actually inaccurate. People search BEFORE they go. In fact, I can tell you the busiest tourism month in Greece is August. I've been there dozens of times and June is actually a pleasant (tho cool-ish) time to visit - no crowds. And the Greeks are still working (trying to make money to spend on their August vacations - when everything shuts down). Perhaps Airline charters would be a more accurate reflection. Altho I do not know where to source that data.

Great site, the UI/design isn't the best though. Improving that will make a huge difference. One of the first things I would do is to make Fahrenheit/Celcius a sitewide setting, like on WeatherSpark.com

This is a great idea, and something I always wished existed - kudos! Seems like there's some great opportunities for SEO and advertising here as well - keep rocking!

I don't know if that Tableau map is going to scale to your desired traffic levels. It's a great enterprise tool, and yeah there are a few publications that leverage it, but my instincts tell me to avoid it on a mid-to-high traffic public page. I could be wrong. It's a difficult graph to replicate under fire, but not so hard that a day or two spent on it wouldn't produce a more performant version.

Very interesting to start. I know for a fact that if you travel to Seattle in Jan/Feb it's really cheap, but of course that's because it will be raining a lot. Rain, for some people is not a real problem, but price is. With that in mind it would be cool to find out when is the best time to travel somewhere the cheapest. That would be really useful for me and probably others.

Cool! I've built similar tool somewhere around 2011. I travel a lot and wanted to have a tool that would help me plan my travels according to specific months. I use it on regular basis since then and it's publicly available at: http://weatherhopper.com

Ooops... looks like the scale on my graph somehow got thrown off. Thanks for the find! I've fixed Cuba manually but will go back through looking for the source of this problem.

Immediately after pageload (without interaction), I get the following on Chrome 64 bit (60.0.3112) on Ubuntu 16.10:

An unexpected error occurred. If you continue to receive this error please contact your Tableau Server Administrator.

Session ID: 7EBED2C1927841ED9575329CE40EB6F7-0:0

Uncaught TypeError: Cannot read property 'refreshImages' of null

Cool idea! I just got back from Austria and was surprised about the rain. Didn't even think to pack a rain coat. Even though I checked the weather beforehand, I figured it was a minor fluke. So hard for a West Coast US person to understand that it rains some places in the summer.

Great idea, mediocre execution, someone else here please get with OP and help him! This is an awesome tool!

Nice one! Just a thought, apart from atmospheric information; you can add Festival into account. Ex. Let's say Holi in India occurs during March, Kite Festival occurs during January. You can add such thing by just combining few places with Wiki Festivals!

Following the feedback from others, a few things seem odd (for Peru):

- I tried it with Lima-Peru and it actually suggested a bad time to visit the city: https://championtraveler.com/dates/best-time-to-visit-lima-p... . The weather on winter is quite humid ( http://www.generaccion.com/noticia/imagenes/grandes/188862-2... )so its not the best time to be there, compared to the summer. - Also, on the mountains the season to avoid is the rain season, because it really rains (jan-feb). - Finally, as others pointed, the choice of locations is a bit odd. There are a few smaller towns but for instance Cuzco or Macchu Picchu are not there

I like this idea a lot, but there is also on thing to consider, Crowds. The best weather also draws the biggest crowds. This drives up food and hotel prices. If you could somehow factor in the crowds and peak season, you could have a really useful tool.

I'm stoked you built this, as I've had the idea and desired the platform for a while. I do a ton of last minute travel and have semi-unlimited options so it's hard to filter down to the best options that offer what I'm looking for.

I bet you could do this by looking at all the Yelp reviews for an area over time. Some combination of average review score and total review quantity would produce a legit metric.

But Yelp and Google's APIs aren't really designed for this kind of use sadly.

Cool I had a similar idea a month ago when looking for a place to go camping. To find the best weather forecast in a range of $x hours drive. Or the best Weather forecast near direct connected airports from my towns airport.

The data can essentially be reduced to a single list on one page because the only thing that is really useful is the range of dates for the best time to travel for each location.

Better yet, add the option to sort the list by dates.

I would love a feature of this where it supports ranges of weeks or for the whole year. I want to use it to find the best place to live based on my preference of yearly weather!

Its funny, it got Vancouver, BC exactly opposite. I'm sure most resident of the notoriously wet city believe July 1-Aug 27 is the best time to visit and not Aug 27-Jul 1

Maybe we can partner up :). I made http://www.averageweather.io for a very similar reason.

Great work! Although it says the best time to visit San Diego is February through November. I guess that's right though; it's great here any time!

It would be great if this included a few more metrics besides weather, like popular tourist times (some people might want to avoid lines and such).

I did something similar as a little weekend hack a few years ago: http://whengo.io

I think recos like this need the local input more than anything else, ex: see a ton of the comments below pointing out issues.

Good start though.

The best times to visit San Diego for ideal weather are February to November

Yeah, that's been my impression, too! :-)

This is cool, where do you get the data? I see its from NOAA, but what kind of data files do you get from them?

This is very cool. Long shot but any chance you can blog about the algorithm used? Or open source the same?

I'm not sure many people would be interested, but I'd have no problem doing a write up with the relevant source. I'll try to get to it next week after fixing all these bugs people are finding.

The New York advice certainly is wrong. Hot humid summer is not New York's best experience.

I had made something similar in 2013 but was not nearly as complete. Good work. Trekweather.com

This is really cool, but why is everything in Fahrenheit? How do I make it default to Celsius?

"The best times to visit Ensenada for ideal weather are January 1st to December 30th"

nice, all year

Time to visit Australia is all year long, but what I hear the best is November or February.

I feel Autumn and Spring are the best times to visit Sydney.

During the Summer you could travel around the southern parts (Tasmania and SA). In the winter months you can spend time in Cairns (anywhere up north really) and its beautiful. You actually don't want to be up north during summer, it is too hot and it's stinger season in the ocean.

How do you get number of Google searches programatically? Is there an API or something?

No, sadly. I had to use a scraper. I think that's part of the reason the tourist volume data was sub-par, so I really need to find a better data source.

This tool doesn’t know about winter destinations. Enter Salt Lake City or Denver.

How does climate change impact your data and site?

Would be nice to be able to select degrees celcius.

How about looking at average flight prices?

cool (°C would be appreciated)

No results found for Osaka :(

Hmm, sorry about that. I'll look into that, but for now the nearby cities of Amagasaki and Sakai are probably very similar.



No Sapporo (or "Hokkaido") either.

Applications are open for YC Winter 2021

Guidelines | FAQ | Support | API | Security | Lists | Bookmarklet | Legal | Apply to YC | Contact