One of my character flaws is that I obsess about weather apps and forecast services. Maybe it's because I don't follow any sports so that only leaves weather to make small talk in elevators. Whatever the reason, I own a lot of weather apps. I fell out of love with Dark Sky predictions several years ago when it felt like they were consistently wrong. I moved to Weather Underground wherever I could, including in CARROT Weather. I'm concerned that the recent API changes at WU are going to mean major changes for the apps I use, so I've been comparing forecasts. Here are a few of my latest obsessive thoughts about weather.
I really like ForecastAdvisor on the web. It provides a simple and easy to read comparison of most of the major services. Many forecasts services can be great for one location and terrible for another. I suspect that there are a lot of fans of Dark Sky around San Francisco. ForecastAdvisor repeatedly shows that Weather Channel and Weather Underground are best where I am.
I also prefer apps like Hello Weather and eWeather HD that allow me to switch between services to get a different angle on a 5 day forecast. I usually care less about app features than I do about forecast precision. But, if you really care about features Versus.com compares a few of the more popular apps.
I also do my own comparisons of weather services. Here's one little slice of high-temp data from a couple of weeks ago. What I care about is how close the predictions are to the reality. You'll see that there's no clear winner but Weather Underground tends to be better than Dark Sky.
Note that these series show a comparison of forecasts between two different days. When I check the weekend weather on Tuesday, I hope that it's still accurate when I check on a Thursday. None of the services really nailed the Sunday weather when I checked on Thursday.
I think a good conclusion for me was that Dark Sky predictions have improved. I continue to collect data and it's holding up. Dark Sky has improved, but Forecast Advisor still ranks it pretty low.
As a side note, if you want to check the actual high temperature in your area, I still prefer good old WolframAlpha.
I don’t believe much of the nutrition “science.” The studies are poorly defined and the results are rushed to a mostly illiterate media. I do, however, believe a broken clock is right two to three times a day depending on daylight savings.
I haven’t found any of the ...
Neil deGrasse Tyson on the rise of crazy flat-earth conspiracy nuts:
I think that the number of people may be the same over time they just now can write a blog that the whole world has access to by a search engine.
I love this YouTube series because I love ...
By way of Kottke, comes this excellent description of the CRISPR technology.
I enjoyed the book Change Agent by Daniel Suarez a lot.1 As with his other books, the basis for the fiction is solid. The futuristic technologies he invents have a firm foundation in actual science. Which is ...
I always enjoy the work over at Information is Beautiful. This post about how scientific theories were received throughout history is very good. I've always assumed astronomy was the most dangerous field for heretics but this data set implies that Medicine will get you more often.
Patrick Soon-Shiong seems like a Perfect fit for the new administration.
If you're interested in what an actually researcher thinks of the Microsoft announcement about intering the oncology research fied, there's no better blog to turn to than In the Pipeline:
Put shortly – and these sorts of stories tend to put actual oncology researchers in a pretty short mood ...
It's a long holiday this weekend in the U.S. Here are a couple of excellent articles about the "Moral Economy" to keep you busy during your travels and parties. What's the moral economy? Well, read them. They aren't just smart, they are well written and often ...
Derek Lowe has some choice comments about the Google biotech Verily.
Here's the quote from their CEO:
. . .We’ve done a lot, to be quite humble about it. Enough to give us great confidence that this is all likely to work. . .
We know that much of this works: we ...
Derek Lowe has some smart things to say about how the humnan genome is portrayed in the media now and how that's going to be changing.
...the steadily increasing power and scope of sequencing technology is allowing us to really get to work in the zone between those two ...
Page 1 / 14