AI-Based Weather Forecasting Has Blind Spots

Traditional weather forecasting models are physics-based and rely on supercomputers. Practically speaking, this means that they start from the basic governing equations (like the Navier-Stokes equations) and use approximations to model aspects of the problem in order to make the physics solvable, given constraints on time, computational power, spatial resolution, and so on.

So-called AI models approach the problem differently, training a model on past weather conditions in order to predict future weather. In some respects, this approach is very successful; AI-based models require less computational infrastructure to run and, in recent years, have greatly improved their predictions of everyday weather.

However, these AI models do poorly when predicting extreme weather events, because their training data contain relatively few examples of these events. They show limited ability to extrapolate their predictions to more extreme events. But these events–like the unprecedented 2021 heatwave in the Pacific Northwest or many of the Category 5 hurricanes we’ve seen in the last decade–are happening increasingly often due to climate change. Those events will keep happening, more frequently, as warming continues. Physics-based models can predict and forecast these events in ways that AI-based models fail to because they are limited by their trained experiences.

Researchers are working to find ways to better equip AI-based models with more physical sense, but, as these models proliferate, it’s important for their users (and those of us using their forecasts) to know what their current weaknesses are. (Image credit: B. McGowan; research credit: Y. Sun et al.; see also S. Nath and T. Palmer; via Gizmodo)

Blue Jewels and Gray Haze

Beginning in early spring, brilliant blue ponds form on Greenland’s ice sheets as meltwater gathers in indentations. This satellite image shows the ice east of Nordenskiöld Glacier, which is the tongue of ice projecting on the left side of the image. The center region of ice is darker, marked by soot, ash, and dirt left behind after previous ice layers have melted. These darker remains make the ice less reflective to sunlight; with less reflectivity, the ice absorbs more sunlight, melting faster. (Image credit: M. Garrison/NASA Earth Observatory)

Predicting Volcanic Eruptions

People have long hoped to reliably predict volcanic eruptions. An automated system at Piton de la Fournaise in France has been doing so since 2014 with an impressive 92% accuracy. The tool, called Jerk, makes its predictions based on real-time measurements of subtle ground movements associated with magma fracturing rock on its way to the surface. Its predictions have ranged from minutes to hours before the start of an eruption.

So far, the team has only tested the system at one volcano, but they are working to install a second version at Mount Etna, where they’ll see whether other volcanoes produce a similar signal ahead of eruption. If so, Jerk could provide valuable warnings in populated areas and give geologists an automated alternative for monitoring remote volcanoes.

To learn more, check out the team’s open access paper and this interview with the team leaders over at Gizmodo. (Image credit: F. Beauducel; research credit: F. Beauducel et al.; via Gizmodo)

Herring Spawn

From mid-February to early May, tiny silvery Pacific herring gather along the shallow coastlines of Vancouver Island off British Columbia, Canada. In these sheltered waters, they spawn; female fish produce sticky eggs and males flood the area with milt, which turns the water a milky turquoise or green. The colors can be so vivid that the spawn is visible to satellites.

Barkley Sound, on the island’s southwestern side, frequently hosts spawning, as its rocky shoreline provides protection and the pockets of lower salinity that the fish favor. After spawning, the fish migrate back to their feeding grounds in deeper, nutrient-rich waters. (Image credit: R. Cutler; via NASA Earth Observatory)

RE: https://mastodon.social/@risujith/116577824196667278

A gentle reminder - the webinar is scheduled for today at 19:30 IST | 16:00 CEST | 07:00 PDT | 10:00 EDT.

#ComplexSystems #webinarinvite #Zoomcodes #IITMadras #IITK #PIK #Nonlineardynamics #Turbulence #BoundaryLayer #FluidDynamics #Bifurcations #FluidMechanics #HeatTransfer #Convection #Scaling

Started a game of dwarf fortress.

My attempt at making a well went poorly. My only three pickaxes are going to be underwater with my three drowned miners. As will my only anvil. I guess I abandon this fort.

The question is, do I try to reclaim it?

#DwarfFortress #FluidDynamics #Oops

Understanding Pollen Dispersal

When the wind blows, trees shift and sway, reconfiguring their shape and their leaves in response. For parts of the year, that flow can also pluck pollen grains off the tree, carrying them on the winds. A new computational simulation models this pollen dispersal from a tree, with the aim of eventually integrating into a tool for urban planners.

Trees are an important component to fighting climate change, especially in cities, because they cool their surroundings in addition to providing fresh oxygen. But urban planners recognize the downsides to trees, too–allergies, anyone?–and, with the right tools, they could maximize the trees’ advantages while minimizing pollen spread for allergy-sufferers. (Image credit: M. Köles; research credit: T. Dbouk et al.; via Physics World)

Setting the Stripes on a Tiger (Cake)

A tiger skin cake forms a distinctive pattern of light and dark patches as it bakes. Its current popularity seems to have expanded outward from China; I found a lot of Swiss-roll-style recipes that use it as an outer wrapper. Here, researchers look at how the wrinkled surface forms. The viscous batter quickly forms a solid skin on its surface, and, as the cake grows, the skin is forced to bend and wrinkle to accommodate the growth. Interestingly, the length-scale of the wrinkling pattern depends on the batter’s depth. For larger stripes, use a thicker layer of batter! (Image credit: K. Koutova et al.)