Nowcasting the next hour of rain - Google DeepMind 31/05/24, 10:15 AM

RESEARCH

Nowcasting the next hour of rain

29 SEPTEMBER 2021

The Nowcasting Team

Share

Our lives are dependent on the weather. At any moment in the UK, according to one
study, one third of the country has talked about the weather in the past hour,
reflecting the importance of weather in daily life. Amongst weather phenomena,
rain is especially important because of its influence on our everyday decisions.
Should I take an umbrella? How should we route vehicles experiencing heavy rain?
What safety measures do we take for outdoor events? Will there be a flood?

Our latest research and state-of-the-art model advances the science of

Precipitation Nowcasting, which is the prediction of rain (and other precipitation
phenomena) within the next 1-2 hours. In a paper written in collaboration with the
Met Office and published in Nature, we directly tackle this important grand
challenge in weather prediction. This collaboration between environmental science
and AI focuses on value for decision-makers, opening up new avenues for the

https://deepmind.google/discover/blog/nowcasting-the-next-hour-of-rain/ Page 1 of 5
Nowcasting the next hour of rain - Google DeepMind 31/05/24, 10:15 AM

nowcasting of rain, and points to the opportunities for AI in supporting our

response to the challenges of decision-making in an environment under constant
change.

Short-term weather predictions

Throughout history, the prediction of weather has held a place of importance for
our communities and countries. Medieval meteorologists began by using the stars
to make predictions. Slowly, tables recording seasons and rain patterns started to
be kept. Centuries later, Lewis Fry imagined a ‘Forecast Factory’ that used
computation and the physical equations of the atmosphere to predict global
weather. In this evolving book of weather prediction, we now add a story on the role
of machine learning for forecasting.

Today’s weather predictions are driven by powerful numerical weather prediction

(NWP) systems. By solving physical equations, NWPs provide essential planet-scale
predictions several days ahead. However, they struggle to generate high-resolution
predictions for short lead times under two hours. Nowcasting fills the performance
gap in this crucial time interval.

Nowcasting is essential for sectors like water management, agriculture, aviation,

emergency planning, and outdoor events. Advances in weather sensing have made
high-resolution radar data–which measures the amount of precipitation at ground
level–available at high frequency (e.g., every 5 mins at 1 km resolution). This
combination of a crucial area where existing methods struggle and the availability
of high-quality data provides the opportunity for machine learning to make its
contributions to nowcasting.

DeepMind About Research Technologies Impact Discover

Overview Blog Events The Podcast Visualising AI

Past 20 mins of observed radar are used to provide probabilistic predictions for the next 90
mins using a Deep Generative Model of Rain (DGMR).

Generative models for nowcasting

We focus on nowcasting rain: predictions up to 2 hours ahead that capture the
amount, timing, and location of rainfall. We use an approach known as generative
modelling to make detailed and plausible predictions of future radar based on past
radar. Conceptually, this is a problem of generating radar movies. With such
methods, we can both accurately capture large-scale events, while also generating
many alternative rain scenarios (known as ensemble predictions), allowing rainfall
uncertainty to be explored. We used radar data from both the UK and the US in our
study results.

We were especially interested in the ability of these models to make predictions on

medium to heavy-rain events, which are the events that most impact people and
the economy, and we show statistically significant improvements in these regimes

https://deepmind.google/discover/blog/nowcasting-the-next-hour-of-rain/ Page 2 of 5
Nowcasting the next hour of rain - Google DeepMind 31/05/24, 10:15 AM

compared to competing methods. Importantly, we conducted a cognitive task

service provider
assessment with more than 50 expert meteorologists at the Met Office, the UK’s
national meteorological service, who rated our new approach as their first choice
in 89% of cases when compared to widely-used nowcasting methods,
demonstrating the ability of our approach to provide insight to real world decision-
makers.

A challenging event in April 2019 over the UK (Target is the observed radar). Our generative
approach (DGMR) captures the circulation, intensity and structure better than an advection
approach (PySTEPS), and more accurately predicts rainfall and motion in the northeast. DGMR
also generates sharp predictions, unlike deterministic deep learning methods (UNet).

https://deepmind.google/discover/blog/nowcasting-the-next-hour-of-rain/ Page 3 of 5
Nowcasting the next hour of rain - Google DeepMind 31/05/24, 10:15 AM

A heavy precipitation event in April 2019 over the eastern US (Target is the observed radar).
The generative approach DGMR balances intensity and extent of precipitation compared to an
advection approach (PySTEPS), the intensities of which are often too high, and does not blur
like deterministic deep learning methods (UNet).

What’s next
By using statistical, economic, and cognitive analyses we were able to demonstrate
a new and competitive approach for precipitation nowcasting from radar. No
method is without limitations, and more work is needed to improve the accuracy of
long-term predictions and accuracy on rare and intense events. Future work will
require us to develop additional ways of assessing performance, and further
specialising these methods for specific real-world applications.

We think this is an exciting area of research and we hope our paper will serve as a
foundation for new work by providing data and verification methods that make it
possible to both provide competitive verification and operational utility. We also
hope this collaboration with the Met Office will promote greater integration of
machine learning and environmental science, and better support decision-making
in our changing climate.

Read the paper Skillful precipitation nowcasting using Deep Generative Models of
Radar in the 30 September 2021 issue of Nature, which contains an extensive
discussion of the model, data and verification approach. You can also explore the
data we used for training and find a pre-trained model for the UK via GitHub.

https://deepmind.google/discover/blog/nowcasting-the-next-hour-of-rain/ Page 4 of 5
Nowcasting the next hour of rain - Google DeepMind 31/05/24, 10:15 AM

Acknowledgements

We are grateful to the Met Office and all our collaborators and advisors for their
input to this work.

Follow us

About Learn more

Sign up for updates on our latest innovation
About Google DeepMind Research

I accept Google's Terms and Conditions and

Responsibility & Safety Technologies
acknowledge that my information will be used in
Education Impact accordance with Google's Privacy Policy.

Careers Discover

Press Sign up

About Google Google products Privacy Terms

https://deepmind.google/discover/blog/nowcasting-the-next-hour-of-rain/ Page 5 of 5