Sonifying the weather forecast

In this project, I create an immersive sonic experience to complement the beautiful sky visualisation of the Norwegian (YR.no) weather forecast app produced by NRK and Met.no.

One of the most attractive features of the sky visualisation is the subtle blending between scenes as the forecast it represents gradually evolves but seems never to repeat. It is what you would see if you looked out of the window and as such is highly accessible. I aspire to similar effect in my sound design where I map the same hourly meterological forecast data to sound. I draw on a combination of synthesised sound elements, and recorded sounds that are manipulated, filtered, triggered and layered in ways that are driven by the data. The blending and mixing of the sound elements as the forecast evolves is a key element of the sound design.

The work presented provides proof of concept, including a suggested protocol for mapping the specific atmospheric elements of the available dataset to sound, a working interactive max8 patch and the demonstration of a potential user experience.

Video demonstration of weather forecast sonification

In this video, my sonification design is coupled to the yr.no weather forecast sky animation. An iphone screen recording shows a user exploring the Edinburgh weather forecast for 48 hours in mid august. The relevant data for the 48 hour period was download from the yr.no forecast, wrangled in python and loaded into max. The soundtrack is generated by sonifying the relevant weather data using my Weather factory max patcher in data driven mode. The timing of the soundtrack was adapted to fit the swipes of the video. For best listening experience please wear headphones.

What we hear

We are listening to the Edinburgh weather forecast for 48 hours from 12:00 on 11th August hours. The listener is facing North and the windows are open.

00:00 - rain throughout the day > moderate rain and moderate breeze from south with a warm soft wind > a reduction to light rain and gentle warm breeze > heavy rain and moderate breeze moving to south west > rain continuing and breeze moving to west > 00:35

00:35 - no precipitation > rain ceases and we hear gentle breeze from west > as night falls and temperature drops wind sounds colder’ (lower frequency) and more directed > 00:57

00:57 - morning birdsong > we are woken at sunrise by a bird chorus > light winds from south become gradually warmer’ > school children play during break time > moderate breeze from south/southwest as a child chatters whilst walking home from school > wind becomes colder’ as temperature drops slightly overnight > 01:52

01:52 - fresh breeze > wind gets warmer and increases in strength > different birds wake us at sunrise > the wind increases to a fresh breeze from south west > 02:05

02:05 - light rain clearing > light rain showers > rain clearing with a moderate to fresh breeze from south west > afternoon traffic > evening live music from the garden drifts on the breeze > 02:41


Programming: Weather factory max patcher

Weather factory, the max patcher (Max8, Cycling 74) written to drive the sound piece includes: wind generator; rain generator; ambience selector; sound mixer and crossfade; sound effects driven by Temperature. It relies on both synthesised sound elements and a suite of field recordings i made in Edinburgh during January - June 2021.


The data

In this example, the relevant data for the 48 hour period was pulled from the yr.no forecast, wrangled in python and loaded into max. The timing of the soundtrack was adapted to fit the swipes of the video.


Mapping weather data to sound

Read more about my design choices for mapping weather data to sound here.


Design

  • Sound design, audio programming and field recording by Karen Mair.
  • Animation from yr.no.
  • Data from met.no.