Singlehanded sailing along the coast of Norway, from the Lysefjord to the Lofoten,
Saturday, January 18, 2020
Friday, January 17, 2020
Google AI model outperforms traditional methods of weather prediction
Weather patterns
Top: Image showing the location of clouds as measured by geosynchronous satellites.
Bottom: Radar image showing the location of rain as measured by Doppler radar stations.
From NeoWin by Ather Fawaz
A couple of weeks back, Google AI used a machine learning model to improve the screening of breast cancer.
Now, the firm has used a convolutional neural network (CNN) in nowcasting precipitation.
In the paper titled "Machine Learning for Precipitation Nowcasting from Radar Images", researchers at Google AI have employed a CNN to give a short-term prediction for precipitation.
And the results seem promising, and according to Google, outperform traditional methods:
This precipitation nowcasting, which focuses on 0-6 hour forecasts, can generate forecasts that have a 1km resolution with a total latency of just 5-10 minutes, including data collection delays, outperforming traditional models, even at these early stages of development.
Unlike traditional methods, which incorporate a priori knowledge of how the atmosphere works, the researchers used what they are calling a 'physics-free' approach that interprets the problem of weather prediction as solely an image-to-image translation problem.
As such, the trained CNN by the team—a U-Net—only approximates atmospheric physics from the training examples provided to it.
For training the U-Net, multispectral satellite images were used.
Data collected over the continental US from the year 2017 to 2019 was used for the initial training.
Specifically, the data was split into chunks of four weeks where the last week was used as the evaluation dataset while the rest of the weeks were used for the training dataset.
In comparison to traditional, venerable nowcasting methods, which include High Resolution Rapid Refresh (HRRR) numerical forecast, an optical flow (OF) algorithm, and the persistence model, Google AI's model outperformed all three.
A visualization of predictions made over the course of roughly one day.
Left: The 1-hour HRRR prediction made at the top of each hour, the limit to how often HRRR provides predictions.
Center: The ground truth, i.e., what we are trying to predict.
Right: The predictions made by our model. Our predictions are every 2 minutes (displayed here every 15 minutes) at roughly 10 times the spatial resolution made by HRRR. Notice that we capture the general motion and general shape of the storm
As can be seen, the quality of our neural network forecast outperforms all three of these models (since the blue line is above all of the other model’s results).
It is important to note, however, that the HRRR model begins to outperform our current results when the prediction horizon reaches roughly 5 to 6 hours.
Precision and recall (PR) curves comparing our results (solid blue line) with: optical flow (OF), the persistence model, and the HRRR 1-hour prediction.
As we do not have direct access to their classifiers, we cannot provide a full PR curve for their results.
Left: Predictions for light rain.
Right: Predictions for moderate rain.
Moreover, the model provides instantaneous predictions.
This is an added advantage because the traditional methods like HRRR harbor a computational latency of 1-3 hours.
This allows the machine learning model to work on fresh data.
Having said that, the numerical model used in HRRR has not entirely been superseded by it.
In contrast, the numerical model used in HRRR can make better long term predictions, in part because it uses a full 3D physical model — cloud formation is harder to observe from 2D images, and so it is harder for ML methods to learn convective processes.
Google envisions that it might be fruitful to combine the two methods, HRRR and the machine learning model for having accurate and quick short-term as well as long-term forecasts.
According to the firm, they are also looking at applying ML directly to 3D observations in the future.
Links :
- Techxplore : Google claims its 'nowcast' short-term weather predictions are more accurate than advanced models
- ClimateChange : Machine Learning for Precipitation Nowcasting from Radar Images
- Phys : Deep learning application able to predict El Niño events up to 18 months in advance
- IBM : New IBM Weather System to Provide Vastly Improved Forecasting Around the World
- GeoGarage blog : Better weather forecasts coming to the ... / Can Artificial Intelligence help build better ... / Using deep learning to forecast ocean waves
Thursday, January 16, 2020
Ocean temperatures hit record high as rate of heating accelerates
The heat in the world’s oceans reached a new record level in 2019,
showing “irrefutable and accelerating” heating of the planet.
Photograph: Modis/Terra/Nasa
From The Guardian by Damian Carrington
Oceans are clearest measure of climate crisis as they absorb 90% of heat trapped by greenhouse gases
The heat in the world’s oceans reached a new record level in 2019, showing “irrefutable and accelerating” heating of the planet.
Earth's global surface temperatures in 2019 ranked second-warmest since 1880, according to independent analyses by NASA and the National Oceanic and Atmospheric Administration (NOAA).
Global temperatures in 2019 were 2 degrees Fahrenheit (1.1 degrees Celsius) warmer than the late 19th century, according to scientists at NASA’s Goddard Institute for Space Studies (GISS) in New York.
2019's temperatures were second only to those of 2016 and continued the planet's long-term warming trend: the five warmest years on the instrumental record have been the five last years.
see NASA
The world’s oceans are the clearest measure of the climate emergency because they absorb more than 90% of the heat trapped by the greenhouse gases emitted by fossil fuel burning, forest destruction and other human activities.
The new analysis shows the past five years are the top five warmest years recorded in the ocean and the past 10 years are also the top 10 years on record.
The amount of heat being added to the oceans is equivalent to every person on the planet running 100 microwave ovens all day and all night.
Hotter oceans lead to more severe storms and disrupt the water cycle, meaning more floods, droughts and wildfires, as well as an inexorable rise in sea level.
Higher temperatures are also harming life in the seas, with the number of marine heatwaves increasing sharply.
The most common measure of global heating is the average surface air temperature, as this is where people live.
But natural climate phenomena such as El Niño events mean this can be quite variable from year to year.
“The oceans are really what tells you how fast the Earth is warming,” said Prof John Abraham at the University of St Thomas, in Minnesota, US, and one of the team behind the new analysis.
“Using the oceans, we see a continued, uninterrupted and accelerating warming rate of planet Earth.
This is dire news.”
“We found that 2019 was not only the warmest year on record, it displayed the largest single-year increase of the entire decade, a sobering reminder that human-caused heating of our planet continues unabated,” said Prof Michael Mann, at Penn State University, US, and another team member.
Oceans are getting hotter due to global heating
The analysis, published in the journal Advances In Atmospheric Sciences, uses ocean data from every available source.
Most data is from the 3,800 free-drifting Argo floats dispersed across the oceans, but also from torpedo-like bathythermographs dropped from ships in the past.
The results show heat increasing at an accelerating rate as greenhouse gases accumulate in the atmosphere.
The rate from 1987 to 2019 is four and a half times faster than that from 1955 to 1986.
The vast majority of oceans regions are showing an increase in thermal energy.
This energy drives bigger storms and more extreme weather, said Abraham: “When the world and the oceans heat up, it changes the way rain falls and evaporates.
There’s a general rule of thumb that drier areas are going to become drier and wetter areas are going to become wetter, and rainfall will happen in bigger downbursts.”
Hotter oceans also expand and melt ice, causing sea levels to rise.
The past 10 years also show the highest sea level measured in records dating back to 1900.
Scientists expect about one metre of sea level rise by the end of the century, enough to displace 150 million people worldwide.
Dan Smale, at the Marine Biological Association in the UK, and not part of the analysis team, said the methods used are state of the art and the data is the best available.
“For me, the take-home message is that the heat content of the upper layers of the global ocean, particularly to 300 metre depth, is rapidly increasing, and will continue to increase as the oceans suck up more heat from the atmosphere,” he said.
“The upper layers of the ocean are vital for marine biodiversity, as they support some of the most productive and rich ecosystems on Earth, and warming of this magnitude will dramatically impact on marine life,” Smale said.
The new analysis assesses the heat in the top 2,000m of the ocean, as that is where most of the data is collected.
It is also where the vast majority of the heat accumulates and where most marine life lives.
The analysis method was developed by researchers at the Chinese Academy of Sciences in Beijing and uses statistical methods to interpolate heat levels in the few places where there was no data, such as under the Arctic ice cap.
An independent analysis of the same data by the US National Oceanographic and Atmospheric Administration shows that same increasing heat trend.
Reliable ocean heat measurements stretch back to the middle of the 20th century.
But Abraham said: “Even before that, we know the oceans were not hotter.”
“The data we have is irrefutable, but we still have hope because humans can still take action,” he said.
“We just haven’t taken meaningful action yet.”
Links :
- The Guardian : Why do record ocean temperatures matter? / Heatwaves sweeping oceans ‘like wildfires’, scientists reveal / UN draft plan sets 2030 target to avert Earth's sixth mass extinction
- CNN : Oceans are warming at the same rate as if five Hiroshima bombs were dropped in every second
- NYTimes : 2019 Was a Record Year for Ocean Temperatures, Data Show
- Time : 2019 Was the Second-Hottest Year Globally on Record, and Ocean Temperatures Are Hotter Than Ever
- BBC : Climate change: Last decade confirmed as warmest on record
- Wired : Our Planet May Be Barreling Toward a Tipping Point
Wednesday, January 15, 2020
Subscribe to:
Posts (Atom)