counters

13 ms·

121.
by counters 2y ago
I'm pretty sure KMUX is fully unobstructed (no beam blockage) at the lowest scan elevation, but I don't have a graphic or source at my fingertips to confirm that. I don't recall any difference in quality between East Bay and
122.
by counters 2y ago
> I didn't live here when Dark Sky was still good, but Dark Sky was incredibly accurate when I lived in the Bay Area. I'm kind of skeptical about that, given that the Bay Area has relatively poor radar coverage. The local NEXRA
123.
by counters 2y ago
Yes; they basically just extrapolated from these "rain blobs" on the visualization as the short-term forecast they provided to users. There are some long-since wiped blog posts that provide a bit more context on how they do a litt
124.
by counters 2y ago
... yeah, but that's just not how it works inside Alphabet. It's not rational but plenty of us would share that this is absolutely the logic that drives day-to-day business decisions across all of Alphabet's business units to
125.
by counters 2y ago
We use a mixture of physical and statistical models to forecast ENSO, including physical models of varying degrees of complexity. A long-standing challenge has been for these models to exhibit reliable skill at lead-times beyond ~6-12 month
126.
by counters 2y ago
There are a few groups working on leveraging observations more directly in the ML forecast models and skipping over the assimilation/analysis step. However, unlike the original ML forecasting problem (which, let's be honest - was
127.
by counters 2y ago
The measurements aloft constrain the entire system - things like vertical profiles of moisture and temperature, as well as the kinematic structure of the atmosphere (e.g. wind profiles) grossly constrain the evolution of the system. Put ano
128.
by counters 2y ago
Parent comment is mis-informed. NOAA doesn't have to buy any data from other government agencies or organizations. It's all open and publicly available. There are challenges around the reliability and quality of data that NOAA doe
129.
by counters 2y ago
The entire slate of commercial acquisitions planned or in progress can be found at [1]. It's pretty anemic; NOAA has spent far less than what folks were hoping they would. I think a major part of this is that the private sector really
130.
by counters 2y ago
These data are already consumed operationally by the major global weather modeling centers - e.g., check out the AMDAR [1] or ACARS programs. There are commercial agreements which restrict third-party access to these data in real-time, but
131.
by counters 2y ago
> NOAA/NWS, for example, is extremely underfunded so if they had to privilege to buy it they probably couldn't come to an agreement to buy it. As a result, they can't use that data to improve the accuracy of alerts/wa
132.
by counters 2y ago
Actually, the paper isn't in press at Nature - it's only published on arXiv. It seems a little bit unusual that Nature would publish this sort of news article for a work that hasn't been peer-reviewed.
133.
by counters 2y ago
The evaluation of the air pollution / chemistry forecasts in the arxiv paper [1] is really threadbare and entirely disconnected from the actual literature of atmospheric chemical transport modeling and air quality forecasting. The high
134.
by counters 2y ago
Why? Most weather and climate datasets - including ERA5 - are highly structured on regular latitude-longitude grids. Even if you were solely doing timeseries analyses for specific locations plucked from this grid, the strength of this sort
135.
by counters 3y ago
"Outperforms" is a stretch; its skill is on par or slightly exceeds the state-of-the-art NWP models currently run today, but you as an end user would not notice the difference between GraphCast and the ECMWF HRES. Furthermore, raw
136.
by counters 3y ago
No, climate models are the same sorts of physics-based simulations as weather models.
137.
by counters 3y ago
This is already SOP at most reputable weather data providers; they consume many different numerical forecasts and use statistical post-processing to choose an optimal blend of the available forecasts based on how different forecasts have ve
138.
by counters 3y ago
There isn't a vector where after-action reports like this could "improve the model." That data is useful for verification, but these systems generally have no learning component to feed the data back into them to improve them
139.
by counters 3y ago
Surprisingly low signal-to-noise ratio for most of the common, creative ways people come up with to detect rain. Windshield wipers on cars are another example. The thing is, even if you did have a super reliable in situ "rain detector&
140.
by counters 3y ago
Well, tracking a rain blob on radar that is 8 minutes from your house is an extraordinarily linear problem, so not surprising they'd have absurdly high P/R for that forecast :)
141.
by counters 3y ago
No clue. They have strong folks on their weather team, too. Not obvious what's gone wrong over there.
142.
by counters 3y ago
Nope. Simple computer vision / optical flow applied to radar image sequences.
143.
by counters 3y ago
Climate change has no impact on weather modeling. The vast majority of weather forecasts derive from physically-based simulations of the atmosphere; the physics of the atmosphere don't suddenly change because the climate is warming. Ho
144.
by counters 3y ago
Dark Sky didn't have "powerful forecasting." They literally just had a simple computer vision app which used optical flow to track blobs and weather radar, and then they extrapolated those blobs forward.
145.
by counters 3y ago
(2) is a big ol' bingo. There was a race towards the bottom line of higher spatial and time resolution over the past 5 years (claims along the lines of "higher resolution means higher accuracy!"), which led to an awful lot of
146.
by counters 3y ago
Do you want a snarky answer or a serious answer? The serious answer is that the way you'd try to figure this out is by combining weather radar, satellite imagery, and a nearby surface observation to try to estimate the current conditio
147.
by counters 3y ago
> Are you saying they want to have racial diversity or SES diversity? Because of structural racism and similar phenomena, they're not fully separable. It's not obvious to me they're even separable in a meaningful way (me
148.
by counters 3y ago
It has to be asked out loud - do you think that the admissions departments at top universities haven't explicitly thought about this? In a post-SFFA vs Harvard world, anecdotes and thought experiments carry very little water. We all kn
149.
by counters 3y ago
Just write your code in JAX and run as usual?
150.
by counters 3y ago
The recent "Neural General Circulation Models" pre-print [1] from Google Research indicates that the team built a spectral dycore from scratch using JAX; in Appendix A they note that it comes in at just over 20,000 lines of code!

More ›