How can a weather app retain users

Weather app comparison: 9 weather apps in the outdoor practical test

Weather apps are a dime a dozen. But which weather app is really suitable for outdoor use? I tested 9 weather apps on my Android phone for you according to outdoor usage criteria.

content

What is important in a weather app

Weather models

What are weather models?

The basis of every weather app is the underlying weather model. They differ, for example, in the regional resolution and the assumptions for calculating the forecast. Ultimately, there are millions and millions of data sets from countless weather stations behind all the different weather models, which are then used for weather forecasts with the help of various algorithms, some of which are adapted to the region.

A distinction is made between global models and local models. While global weather models have a mesh size of 10-50km, local models can also predict smaller areas with a mesh size of 1-15km (source). You shouldn't be confused by the term “local weather model”. So with "local" here is meant more regionally such as Europe. But here, too, the influencing factors on the weather, e.g. due to the topography, are different than e.g. in North America. For example, the American GFS global model is often applicable to North America, but in Europe it provides poorer forecasts than local weather models. (Source)

Depending on the weather app, several or individual models are used. It can therefore be worthwhile to look specifically for local websites (e.g. the Wetterochs for the Regnitz catchment area) and local apps (e.g. Veður for trekking in Iceland, MeteoSwiss for Switzerland) before a tour.

Weather model comparison

Windy has published a nice overview map of the coverage of various weather models:

Weather models with their coverage (© Windy.com)

global weather models:

  • ECMWF / EZMW: Abbreviation for "European Center for Medium-Range Weather Forecasts" (resolution 9km)
  • GFS: Global Forecast System (American weather model, resolution 22km)
  • HRES: (operator: ECMWF, resolution 9km)

local weather models

  • ICON: Icosahedral Nonhydrostatic Model (operator: DWD, resolution 7km)
  • NEMS: Nonhydrostatic Meso-Scale Modeling (operator: Meteoblue, resolution 4 (Central Europe) - 12 (Europe) - 30 (Global) km)
  • AROME: Application of Research to Operations at MEsoscale (resolution 2.5 km; resolution in Germany, France and the Alps 1.25 km)
  • AROME-Arctic: adapted AROME model for the highest north of Europe (operator: Meteorological Institute Norway, resolution 2.5km)
  • UKMO: United Kingdom Met Office (resolution 1.5km)
  • HIRLAM: High Resolution Limited Area Model (origin in Norway, resolution 2.5km)
  • NAM: North American Mesoscale Forecast System
  • MEPS: MetCoOp EPS, AROME modification for Scandinavia (resolution 2.5km)

(Resolution: for the short-term forecast up to approx. 36h, longer-term forecasts sometimes more coarse-meshed)

But do the weather models even differ? You can see that very easily if you use a page that compares the forecasts of different weather models in multi-models:

Weather model comparison from Windy (screenshot)

As you can see in this example (here: Muggendorf for paddling on the Wiesent), the temperatures between the different models only differ by approx. 1-2 ° C in the short-term forecast. Only with longer forecast periods does the range of fluctuation increase a little. The deviations in precipitation and wind are more relevant. Even if in this example it cannot spoil the tour with such a low rainfall, you can still see the difference between the weather models. Good chances of getting the best possible forecast by selecting the best local model for a trekking tour and not relying on global weather models that are too rough.

In practice, however, even with the best local selection, no algorithm is fully developed to 100% perfection. It is not without reason that the Kachemannwetter weather channel writes ...

The thing with apps or online symbols is always such a thing. You have to do a little bit to predict your location think for yourself and then also becomes the perfect prediction to have.

Therefore, it makes sense in any case not only to concentrate on the icons but rather on the absolute values, trend developments as well as radar and satellite images for weather forecasting. Otherwise you will end up missing a nice tour just because the colorful icons for your own local tour area are far too defensive.

The perfect weather forecast for the day

In my opinion, there are two practicable ways to get the best possible weather forecast:

  1. Use weather forecast from the current best weather model for the day
    1. View the app with different weather models (e.g. Windy)
    2. Compare the current weather with the forecast displayed by the various weather models
    3. for the daily forecast choose the model that best reflects the current weather
  2. Derive the weather yourself from radar and satellite data
    1. select a suitable app that provides the relevant data
    2. view the course over time in radar and satellite images
    3. Interpolate the further course with common sense - after all, a rain front does not suddenly turn the direction at 90 °

Actually not that difficult, is it?

Weather app functions

In this section you will learn what I look for in a weather app for trekking and outdoor use in addition to the standard forecasts (temperature, amount of precipitation, probability of precipitation, sunrise / sunset times). Ultimately, I want a weather app to get the best possible forecast on the one hand and to be able to draw my own conclusions from the data on the other. For this I pay attention to:

  • Weather model (s)
  • Multi-model
  • Update frequency (as high as possible, at least 4 times a day for critical weather conditions)
  • Resolution of the forecast (if possible every hour)
  • (Rain) radar image with animation
  • Satellite images with animation
  • Wind speed and wind direction
  • Meteograms
  • Freedom of advertising (focus on content, reduction of data consumption)
  • little data consumption (important in areas with poor network coverage)
  • good usability (simple, understandable presentation and operation)

Practical additional functions:

  • Allergy tracker
  • Storage of favorites
  • Weather data according to altitude (for mountain tours or aviation sports)
  • Snow depths and new snow forecast

The weather apps in comparison and the practical tests were allowed to face precisely these criteria. And then I experienced a few surprises myself.

Weather apps overview

MeteoBlue

For me, Meteoblue was the first weather app that I tried and used for a long time. The weather data based on the in-house weather model NEMS (resolution 4km) are visually very nice, but almost overloaded, processed and they provide comprehensive meteograms. For me, however, the visualization is often too conservative. Unfortunately, I didn't do some tours earlier because of the weather forecast, even though the weather would have been good enough for a hike. The app also provides a large amount of data, images and maps that also enable your own weather forecast

Meteoblue: forecast, meteogram, weather map

Weather model (s)> 25, including Arome, HIRLAM, ICON, NEMS, GFS, COSMO
Multi-model
Selection of the weather model-
UpdateTwice a day
Resolution of the predictionhourly
Radar images with animation✔ (rough map grid)
Satellite images with animation✔ (rough map grid)
Wind data✔ (no map display)
Meteograms
Advertising freedom- (can be switched off via in-app purchase)
Storage of favorites- (for "last places")
Weather data according to altitude-
Snow depth / fresh snow- / -

Ultimately, Meteoblue is an app with a very extensive database. The basic functions such as hourly displays, meteograms, multi-models for temperature and precipitation are convincing. You only have to make compromises for the almost too colorful design, the rather rough maps in the free version and the lack of option to choose the weather model yourself.

YR.no

I came to appreciate and love YR on my tour through the Hardangervidda. It has a simple structure, requires little data volume and was also very precise. Operated by Norway, it offers data updated four times a day for Norway, Sweden, Denmark and Finland while the worldwide data is updated twice a day.

YR.no: weather forecast, precipitation and temperature, wind and pressure forecast

Weather model (s)AROME MetCoOp, AROME Arctic, HRES
Multi-model-
Selection of the weather model-
Update4x daily for Norway, Sweden, Finland, Denmark
Twice a day worldwide
Resolution of the predictionhourly
Radar images with animation-
Satellite images with animation-
Wind data
Meteograms
Advertising freedom
Storage of favorites
Weather data according to altitude-
Snow depth / fresh snow- / -

The simple, ad-free app is extremely fast in practice and uses little data. The simple representations with hourly resolution are easy to understand even for laypeople.

The weather models are a special feature here. The operator of the app, the Norwegian Meteorological Institute, operated and further developed special local weather models in cooperation with neighboring countries. In addition, the predictions are then refined with your own corrections based on empirical values:

local weather models for Scandinavia (© YR.no)

So it's no wonder that the app has absolutely proven itself in practical use in Scandinavia such as Norway (e.g. Jotunheimen, Hardangervidda) or Sweden.

WeatherPro

I came across WeatherPro through a recommendation from a friend. At first glance, the app also likes the tidy layout and the range of functions. Unfortunately, a lot of data is already hidden behind a payment barrier. For example, most of the meteograms (only the wind meteogram is free) and the hourly forecasts are only available in the Pro version.

WeatherPro: forecast, rain radar, meteogram

Weather model (s)in-house multi-model MOS based on ECMWF, EPS, UKMO, NCEP and own data (unfortunately no information on the resolution)
Multi-model-
Selection of the weather model-
Updateseveral times a day (no precise details)
Resolution of the predictionEvery 3 hours in the free version
every hour in the pro version
Radar images with animation✔ (15min grid, deep zoom)
Satellite images with animation-
Wind data
Meteograms✔ (in the free version only wind data)
Advertising freedom✔ (but some functions only with Pro version)
Storage of favorites-
Weather data according to altitude-
Snow depth / fresh snow- / -

In the test with data for Esporles on the GR221 in Mallorca, the free version unfortunately could not be started several times. When it started, it simply remained on the start screen without any error message under the current Android operating system (9). If that happened on tour you would be without a weather app. For this reason alone, the Weather-Pro app is ruled out for my touring use. Fortunately, it remains the only app with a total failure in this comparison and if it runs stably with you, nothing should prevent its use.

DWD WarnWetter

The DWD WarnWetter - App does what it says in its name. It warns of dangerous weather situations and natural hazards such as wind, heat or avalanches or storm surges and, if desired, also via push message.

The app only contains general weather data such as rainfall, temperature models or wind forecasts after an in-app purchase. The few data you then get are very detailed. For example, the radar animations are available in a 15-minute grid. Geographically, the app remains limited to Germany, even in the purchase version. This means that the app is ultimately only an option for trekking in Germany.

AccuWeather

The worldwide widespread weather app can boast 50 million + downloads. And it actually has a large amount of data in the hourly forecast when you click on the details. In addition to the standard weather forecasts, it also offers a pollen forecast, real-time forecasts, severe weather warnings, snowfall and dew point. You look for wind data in vain and the radar map is not really self-explanatory to me with its different views. Overall, I find AccuWeather a bit old-fashioned in terms of layout and user guidance.

The app is financed by advertising that is only indicated by a placeholder in my test, which can also be completely switched off through in-app purchases.

AccuWeather: Realtime weather, hourly forecast, radar map with weather warning

Weather model (s)n / a
Multi-model-
Selection of the weather model-
Updaten / a
Resolution of the predictionhourly
Radar images with animation✔ (cumbersome operation)
Satellite images with animation-
Wind data(✔) (only wind speed, no map display, no direction)
Meteograms-
Advertising freedom- (can be deactivated with in-app purchase)
Storage of favorites- (for "last places")
Weather data according to altitude-
Snow depth / fresh snow- / -

Unfortunately, Akkuweather is very economical when it comes to publishing the underlying data. Without knowing which weather models are behind the forecasts and how up-to-date they are, in the end the forecasts are just crystal balls for the user.

Windy

Windy was born out of a passion for the outdoors. And you can tell. In the comparison field, Windy has by far the largest range of functions - even with the free forecast in the 3h grid. Hourly forecasts, four updates per day and no advertising are available with a monthly or annual subscription (approx. € 20).

The range of functions is convincing, but see for yourself:

Windy: basic view, map, meteogram

Weather model (s)AROME, NEMS, ICON, ECMWF, GFS, NAM
Multi-model
Selection of the weather model
Update2-4 times a day (depending on the weather model)
Resolution of the predictionhourly
Radar images with animation
Satellite images with animation
Wind data
Meteograms
Advertising freedom
Storage of favorites
Weather data according to altitude
Snow depth / fresh snow✔ / ✔

6 different weather models are available depending on the region, radar and satellite data with animation allow your own forecasts and the possible data fields are really extensive (including snow depths, humidity, visibility, wave heights, snow line and much more). With all the functionality, the app just has to be careful that it remains easy enough to use in the future. So far, it has absolutely convinced me in practical use.

Wind finder

The Windfinder app is especially interesting for sea kayakers. I can only speak for lakes, not for the "real" lake. Here the app offers nicely prepared wind data for the paddler. But I don't use them enough myself for a final judgment. For a first look whether a tour on inland lakes is worthwhile, I mainly use it because of. the map function. On the other hand, the Windy app offers that and much more besides. Therefore, in my opinion, Windfinder is not worthwhile, especially since it also contains quite intrusive advertising.

Weather Online

The weather app WetterOnline is one of the apps with the most downloads in my comparison. It has already been installed more than 10 million times. But even the first impression disqualifies the app for outdoor use for me. It is completely overloaded with advertising, which is why I also forego screenshots at this point. In areas with poor network coverage, this simply costs far too much data volume, time and thus electricity. You can switch off advertising with a subscription, but the range of functions is simply too small for that, even if there is a nice function with pollen count data that not all apps have to offer.

RainRadar

The name says it all. The "RegenRadar" app from WetterOnline offers rain radar for Germany, Austria and Switzerland for 90 minutes in the past or for the future. A nice gadget, but a much too short observation period for touring. Who wants to watch the RegenRadar all the time while on tour?

Weather app comparison

In the final weather app comparison, only a few of the apps briefly presented before made it. For example, all apps that do not provide transparency about the data models in the background or that only provide very specific data have been eliminated. Finally, I want to find out which weather app is best for trekking in Europe as an all-rounder.

The 4 remaining apps in direct comparison:

(Legend: ✔ - available, ✔✔ - good, ✔✔✔ - test winner)

Conclusion: A good weather app doesn't need advertising

The single weather app that provides the ultimate weather forecast for your own route for all tours in Europe simply does not exist in my opinion.Rather, it depends on which weather model fits the tour area, what you expect from the app and how far you are ready and able to predict the weather yourself to a certain extent. However, there is one insight:

A good weather app doesn't need advertising and doesn't have to cost anything!

My personal favorites are both free:

  • YR.no for trekking tours in Scandinavia. The app's forecast is sufficient and updated frequently enough, even without detailed knowledge, to be able to adequately estimate the course of the day with minimal data volume. In addition, the app has a simple, uncompromising structure, is ad-free and can be understood by any layperson.
  • Windy for all tours outside Scandinavia in Europe (whether trekking, hut tours, paddling or bikepacking). Starting with the selection of the right weather model for the current tour area, through the countless functions to the radar and satellite images, the app simply offers everything we need outdoors. The app is free of charge in a 3-hour grid, hourly data is only available with a subscription.

What I wouldn't do with any weather app: share my location or share usage data. On the one hand, the switched-on GPS module and, above all, the permanent network operation of the smartphone costs electricity, on the other hand you never know what will really happen to the data in the end.

You might also be interested in: