“Which witch knew whether the weather …

… was going to change.” Ever heard of that spelling challenge?

Spelling aside, how do we know whether the weather is going to change? Now that WeatherSignal on iPhone 6 and 6+ can collect barometric pressure, and given the foggy, rainy weather we’ve been experiencing in London recently, it’s a great time to explore what you can discover about weather and forecasting by measuring barometric (or atmospheric) pressure.

How well can we understand the weather with only barometric pressure? In their paper “Surface pressure observations from smartphones” (which uses WeatherSignal data!), Clifford Mass and Luke Madaus explain that pressure may be one of the most valuable weather measurements, as it is not easily affected by other atmospheric factors, and because it reflects the deep, 3D structure of the atmosphere.

(Of course, a fun project might be to attach an iPhone with WeatherSignal to a drone and measure pressure at different heights! That way, you could build your own model of surface pressure and how it relates to the pressure throughout the air column.)

Hobbyist drone Flying over field

Hobbyist drone flying over field; Photo Credit: Michael Khor

Okay, so we’ll use a robotic flying drone or a model airplane to take our pressure measurements! Well, maybe. In any case, what should we be looking at? What are the components of a forecast based on pressure?

I can think of a few variables that would be of interest (and welcome any comments):

  • The absolute value of pressure (WeatherSignal for iPhone shows you the absolute pressure)
  • The absolute pressure at sea-level, which is essentially absolute pressure normalised for altitude (if you sign up for the WeatherSignal on iOS beta-testing programme, you can get both the absolute pressure and the adjusted sea-level pressure).

  • The rate of change in pressure - how quickly pressure changes is reflected in the intensity of weather changes. This article give a nice overview on how different rates of change in pressure relate to weather.
  • How much the pressure changed - hurricanes are often accompanied by pressure drops of 30 to 70 mb (millibars).
  • The direction of weather changes – going from low to high likely means better weather, whilst if the pressure drops, get your umbrella! (or go inside!)

It would also be interesting to see how the pressure readings from WeatherSignal correlate with the forecast of your local meteorological office. In looking for pressure maps for London (provided by the UK’s MetOffice), I found some great resources to help you read weather maps of all sorts:

  1. UK’s Met Office Surface Pressure Maps
  2. Factsheet on interpreting weather charts

All of this is really interesting, but I feel that we can push the boundaries even further in terms of making connections between weather and other, non-weather phenomena. As it is, we’re still in the “Weathersphere”! Looking for fun or interesting connections, following are a few that I found.

Barometric pressure has an effect on water chemistry

According to the Environmental Monitor, pressure has an effect on water chemistry, specifically the amount of gas that can dissolve in water increases with pressure. This article on pressure points out that this applies to oxygen, which makes me think about the health of ecosystems and how changes in pressure affects animal populations.

Pressure as an indicator for man-made climate change

A new study led by Stanford climate scientist Noah Diffenbaugh finds that California’s drought, which has been marked by exceedingly high atmospheric pressure, is much more likely to have happened in conditions with modern concentrations of greenhouse gases. What if these findings also indicate that pressure could be an indicator of air quality? It would be interesting to see if there is a significant difference in pressure readings in cities versus rural areas (on top of from normal atmospheric changes).

Pressure as an indicator of animal migrations

Although it is understandable that weather affects animal migrations, and that pressure is directly related to the weather conditions, using pressure for understanding ecology and animal activity has interesting potential. What phenomena can we observe by superimposing pressure data by what we observe in our parks or gardens?

What connections do you observe? What analysis would you use barometric pressure for? Just starting out with your explorations of the Weathersphere? Let us know how WeatherSignal works for you!

Pressure Level: Sea

One of the things that we are often asked is ‘why does the reading from the WeatherSignal app differ from the reported pressure on the weather forecast?’ Ok, maybe it’s not all that often, but it still gets asked and is a very good question.

To understand the difference between the barometric pressure reported by WeatherSignal and the pressure reported on the forecast, we have to first understand a concept called Sea Level Pressure – which is (as you might have guessed) the pressure at sea level.

Pressure is essentially a measure of the mass of air above you – and so the higher you go the lower the pressure will be in your location. To get around this and to make showing pressure across countries and regions comparable, pressure readings are transformed into sea-level pressure readings – so that the isobars showing fronts of pressure don’t end up simply reflecting changes in the elevation of the earth. Put simply, WeatherSignal simply shows you the raw pressure reading taken at your location – it does not account for altitude. If you live somewhere high up (Hi WeatherSignal users in Denver!) then WeatherSignal will consistently give you a different pressure reading to the forecast. For those of you who live by the sea or in low-lying cities (like us in London) then WeatherSignal will return a pressure value almost identical to the forecast.

What is most interesting about pressure as an atmospheric variable capable of telling you about the weather is, however, not the absolute value – but rather the rate of change. Pressure change pretty much tells you the speed at which the weather is changing (and also gives you information about wind – which is essentially air flowing between areas with a pressure differential). So, even if WeatherSignal is not reporting an accurate absolute value, it can still tell you that the weather is improving or getting worse!

For more information on how to calculate sea level pressure from a raw pressure this [pdf] page may prove instructive.

WeatherSignal for iPhone

Today marks an exciting step forward for the WeatherSignal project – we’re finally able to expand the scope of the project to include iPhone users, as the iPhones 6 and 6+ contain pressure sensors that are accessible by developers. The new WeatherSignal app for iOS is a fully functioning iPhone barometer that also lets users contribute to our existing weather crowdsourcing project – and lets the user see their own, and others’, readings in-app.

WeatherSignal3   WeatherSignal4  WeatherSignal1

The existing Android app has been downloaded over 270,000 times, with 50,000 active users worldwide and we’re hopeful that the iOS version will help us expand our network to make the world’s most comprehensive mobile meteorology network. Currently, researchers in several countries are using the data collected from the WeatherSignal network to study how it can be used to improve weather forecasting and better understand urban climate systems. The demand for an iOS app (impossible before the 6 was released) can be shown by the enormous demand for the StormTag Kickstarter. StormTag, a key-ring pressure and temperature sensor that will sync up with the WeatherSignal app on both Android and iOS, aimed to raise $17,500 but ended up smashing that target and raising $112,000 in total!

Download WeatherSignal for iOS here.

Update: We also wrote a Vanilla Ice parody blog post announcing the launch (don’t ask us why)

Teaming up to build a $20 Weather Station

We love smartphones, especially ones with a lot of sensors, currently they’re the source for all the data here on our website and on the maps in our Android app.

But there’s a problem, a lot of phones don’t have the most valuable meteorological sensors. The pressure sensor – the most crucial sensor for the this crowd-sourcing project – is on only about 25% of Android devices. And 0% of iOS devices. Ambient temperature sensors are on only 1/10 Android devices, though we are able to get round this by using battery temperature sensors – but it means we need hundreds of phones in an area to generate a clean temperature reading.

Today we’re announcing a project to make it possible for iPhone users and Android devices without pressure/temperature sensors to contribute data. We’re teaming up with Hex3.co to make WeatherSignal the companion app to StormTag. StormTag is a piece of hardware that fits on a keyring, never needs charging, and wirelessly sends pressure/temperature data to an iOS or Android WeatherSignal app. And here’s the kicker, it costs $20.

Here’s Jon Atherton form Hex3.co introducing the StormTag on KickStarter:

Visualization of the week: The world wind map

In order to make the WeatherSignal blog more exciting, and mainly because we’re dataviz nerds, we’re now going to be doing a weekly post on our favourite data visualizations from around the internet – in order to justify quite how much time we spend talking about them. Our inaugural data-visualisation for 2014 is one that ties in extremely well with what we’re doing here at WeatherSignal – presenting live weather data in new and interesting ways. Mapping (almost) live wind patterns across the world ties in perfectly with our project to make pressure readings more granular – and is a particularly beautiful way to visualize the global weather system.

world wind gif 2

This visualization by Cameron Beccario gives a beautiful perspective on the atmospheric topography of the earth. Make sure you check out the site itself, especially the different ways you can visualize the earth’s surface. Based on data collected by the NCEP/National Weather Service/NOAA this visualization beautifully renders the movement of wind across the world’s surface, with an easily comprehensible colour-code to denote speed. What makes this visualization so successful is the way it demonstrates the inter-connected nature of the global weather system, illustrating the sloughs and troughs of pressure that drive the air’s movement.

By emphasizing the fact that weather is a global system the earth wind map shows exactly why weather forecasting is such a difficult process – with so many different contributing factors producing the specific atmospheric conditions we experience. This is something we hope to be able to improve on using the data we collect from WeatherSignal, more granular pressure readings should lead to more accurate short term storm forecasting. For now, though, just enjoy playing around with this great interactive map. Last I checked there was a storm whirling happily off the coast of Madagascar, if only this map also showed the recklessly fluttering butterfly that caused it.

Screen Shot 2014-01-02 at 14.12.38


Our Academic Partners

WeatherSignal has now been up-and-running for just over eight months, and in that time we’ve really seen the project take off. We’ve had around 150,000 total downloads, with almost 45,000 users regularly contributing data from across the world. In only eight months we’ve managed to build the world’s first crowdsourced mobile weather network.

To get an idea of what an average day’s worth of data looks like, here’s a map of all contributions to the WeatherSignal project from the last 24 hours.


Our belief in the role WeatherSignal can play in gathering data for meteorological purposes is founded in our commitment to the idea that mobile phones are giving us the ability to completely transform the way in which we are able to understand the world. There are numerous examples of mobile sensor networks being used in innovative ways, and, with WeatherSignal and OpenSignal we are at the forefront of this technological revolution.

One of the questions we’re often asked is ‘what do you do with the data?’ The answer to that question is really twofold – we visualize it on this site, and we provide it as an hourly data dump to our academic partners, with the intention of helping them prove the value of smartphone-collected meteorological data. Below, I briefly summarize four of the most exciting projects using WeatherSignal data.

Dr. Art Overeem

Earlier this year we published a paper in conjunction with the Royal Meteorological society in the Netherlands, which used cell phone battery temperature readings to approximate ambient temperature. Art is now using the temperature readings collected from WeatherSignal smartphones to examine the historical correlations in other cities.

Birmingham University

The main challenge associated with finding a use for the WeatherSignal data is proving the validity of the readings that can be taken from smartphones. In order to do this we need to compare a high volume of smartphone readings against a similarly dense network of designated weather stations. Birmingham University Climate Lab (BUCL) have a network of over a hundred weather stations set up across Birmingham, and we are excited to have the opportunity to use this network to prove that smartphone atmospheric readings can be both useful and valuable. The main emphasis of this approach is using smartphone readings to study urban climate, and we are very excited to see how this project turns out.

WOW (Met office)

The WOW initiative at the Met office is an exciting step forward towards citizen contribution to weather data collection. WOW allows owners of personal weather stations (which have increasingly come down in price) to share their data on a live map hosted on the Met Office website. We are in talks with the Met office to add a WeatherSignal overlay to this map, and the Met Office are very excited about exploring the potential for smartphone-collected data.

Screen Shot 2013-12-04 at 13.40.10

A snapshot of WOW Data

University of Washington

One of the challenges associated with using smartphone readings for general weather forecasting is the fact that the weather must be understood as an expression of atmospheric conditions. The Atmosphere is three dimensional, and therefore surface readings can only get you so far – which is why much of the work of weather forecasting is done by satellite.  That is not to say that surface readings aren’t useful, they most certainly are, but they do have their limitations. but it is pressure readings that offer the clearest route towards incorporating smartphones into mainstream weather forecasting. Clifford Mass, at the University of Washington, is currently making use of the WeatherSignal data stream to investigate the effectiveness of using smartphone barometers for this very purpose. We are very optimistic that Cliff will be able to show that a network of smartphone barometers will be able to improve both the effectiveness and granularity of  weather now-casting and forecasting.

We’re always looking for more partners to help us explore the potential for the WeatherSignal network, so please don’t hesitate to get in touch via the contact form. We’d love to hear from you!