Now that NASA has released their data updated through July, we know that in that data set, this July was the hottest July on record with a temperature anomaly of 0.75 deg.C, i.e. it was 0.75 deg.C above “climatology” (which is what’s usual for the given month). It’s not the hottest temperature anomaly in the data set, however; that record still belongs to January 2007, at 0.96 deg.C above climatology.
Yet it does seem that this July, while not the hottest temperature anomaly on record, is the hottest month on record.
Fires which burn 1,000 acres or more are considered “large.” That’s well over a square mile (1.5625 to be precise). Imagine a wildfire consuming an entire square mile — it’s huge, clearly a devastating conflagration.
So when I learned that the “Rocky fire” in California had burned through more than 100 square miles, I knew it was huge beyond huge, the biggest fire I’d ever heard of. It’s the kind of raging inferno that costs so much money, property, even lives at times, that takes so many resources to fight, it’s more than just a drain on resources, it’s a disaster, at least for those who live in the vicinity.
Note: See the update at the end of the post.
Since I’ve devised a new way to estimate the survival function, I wrote some code (in R) to execute it. The program is in its early stages of development, so there’s a lot of work still to be done. But one of the ways to learn about it is to let people play with it. So, I’ve decided to share my code so others can have some of the fun.
In the last post I mentioned that I’ve come up with a new way to estimate the survival function. I first got the idea years ago while studying survival analysis. But recently I became aware just how troublesome it is to get good estimates of confidence limits, so I elaborated on the idea and came up with what I think is the answer.
Of course I could be wrong. And of course it’s possible somebody else already thought of it, and it’s published somewhere I’m not aware of. That’s the way science works. But I have submitted a manuscript for publication, so I should find out before too long. In the meantime, I’ll share it here.
The topic: it’s probably not what you think.
This post has nothing to do with climate change. It’s about the survival function, which finds application in many fields, most notably medical research (where survival is considered important) and failure testing. I’ve come up with a new way to estimate it from observed data, and most important, what I believe is a much-improved way to estimate confidence limits.
(Note: this is chapter 1 of my book “Noise: Lies, Damned Lies, and Denial of Global Warming.” It’s not one of my usual technical-type posts, it’s an attempt to illustrate for the general reader how statistics can be misused to deny global warming. The data used are several years out of date, but the point is still quite valid, and at the end I’ve added a graph of more current temperature data. Feel free to refer your friends to this.)
Most of us have heard of the game called “Russian Roulette.” A revolver (usually with 6 chambers) is loaded with only one bullet; five chambers are harmless but one is lethal. The cylinder is spun so the location of the live but deadly cartridge is randomized, then the players take turns putting the gun to their head and pulling the trigger. Sometimes the cylinder is spun again before each turn,
sometimes not. Whoever is unlucky enough to happen on the “live” chamber gets a bullet to the head and almost certain death; the other players are the “winners.”