Author Topic: Week 3: Chapters 4 & 5  (Read 2887 times)  

Offline Eltanin Publishing

  • Status: Scheherazade
  • *****
  • Posts: 1266
  • Charlotte, VT
  • Cathy Ryan - Editor-in-Chief
    • View Profile
    • Eltanin Publishing
Week 3: Chapters 4 & 5
« on: November 26, 2012, 07:01:46 am »
I'm going to try something different and start this thread without looking at the book for what I highlighted - I'll start with what has stuck in my head (and then will add more later after looking at my notes/highlights).

So, chapter 4 was about weather forecasting, and chapter 5 was about earthquakes. The results of years of computer advancements were night & day in these two disciplines - much improvement in weather forecasting, and essentially no improvement in earthquake forecasting. Or rather, based on the USGS's definitions of prediction and forecasting that Silver provided, I think the proper way of saying it is that there has been some progress in forecasting (likelyhood of an earthquake in certain areas, where the dangerous areas are, and the probability of minor vs. major quakes), but no progress in prediction - being able to say a serious quake will hit a certain area tomorrow. Silver points out that different disciplines may use different definitions of prediction and forecasting.

I found it very interesting how local weather forecasters are less accurate than the National Weather Service, because they purposely fudge things. They have overly "wet" forecasts because people get more annoyed when it rains when rain wasn't predicted, than if it is sunny when rain was predicted. It's hypothesized that this contributed to people not taking Katrina seriously - the local weathermen tended to "cry wolf" and just in general were not accurate, so people don't listen.

I'm amazed (though I already knew this before reading this chapter) that we have no way of predicting earthquakes (other than the slightly increased risk after a "swarm" of smaller quakes). On one hand, it makes sense - we can't tell what's going on many miles below the surface. But on the other hand, we know what's going on in other galaxies, or on a microscopic level inside human cells, yet we can't figure out our planet.

The weather chapter reinforced the notion in my head that when a weather forecaster says 40% chance of rain, it means that of all they times they predict a 40% chance of rain, it will rain about 40% of those times. On any given day, it will rain or it won't - there's no 40% about it.
 

KBoards.com

  • Advertisement
  • ***

    Online Ann in Arlington

    • Global Moderator
    • Status: Shakespeare
    • *****
    • Posts: 67017
    • Not really a moderator. :)
      • View Profile
    Re: Week 3: Chapters 4 & 5
    « Reply #1 on: November 26, 2012, 06:57:14 pm »
    I wasn't particularly surprised by either the weather or earthquake chapters. . . .more or less confirmed what I'd often suspected. . . .especially about the local weather guys generally going conservative in terms of not being willing to say it WON'T rain on a given day.  We have one guy around here, though, who's always been really clear about stuff. . . .actually, there are several who I think tell it like it is.  But the one guy is Bob Ryan and he'll very infrequently give a 'Bob Ryan Guarantee' about rain or snow or sunshine. . . . .if he says that, you can really on it with pretty much 100% assurance because he doesn't do it very often.

    Having lived in Hawai'i and California, the lack of reliability of earthquake prediction or forecasting wasn't any big surprise either. ::)
    My Kindles
    Hermoine's Handbag (Voyage)
    Ed's (Voyage Refurb)
    Bedtime Reading (Oasis 9)
    Godric's Hollow (Basic 7)
    My Tardis (PW 10)

    Offline Eltanin Publishing

    • Status: Scheherazade
    • *****
    • Posts: 1266
    • Charlotte, VT
    • Cathy Ryan - Editor-in-Chief
      • View Profile
      • Eltanin Publishing
    Re: Week 3: Chapters 4 & 5
    « Reply #2 on: November 28, 2012, 09:57:07 am »
    I was just pondering something regarding "wet" forecasts - forecasters covering their behind by slightly increasing the chance of rain, since people are more likely to get upset about rain when sun was predicted, compared to an unexpected nice day.

    I was wondering what about in situations where opinions differ on what is "good." For instance, my husband I, and many other Vermonters, love snow. We love downhill and cross-country skiing, sledding, snowshoeing, and we think it's pretty, and we just think it should snow in Vermont in the winter. Others (even here in VT) don't like snow. Some newscasters will say things like, "We dodged a bullet on that one!" when a storm passes us by. I would say that more than half of Vermonters root for snow (because even if you don't ski, you know the tourist industry depends upon it). So I wonder if local forecasters are more likely to lean "dry" - saying there's a less chance of snow than what the National Weather Service (NWS) says, because they don't want skiers to get annoyed at them if it doesn't snow. On the other hand, it's a safety issue since snow plows, school cancellations, etc. depend upon an accurate forecast.

    I also wonder if local forecasters still lean "wet" when they are in an agricultural area in the midst of a drought.
     

    Online Ann in Arlington

    • Global Moderator
    • Status: Shakespeare
    • *****
    • Posts: 67017
    • Not really a moderator. :)
      • View Profile
    Re: Week 3: Chapters 4 & 5
    « Reply #3 on: November 28, 2012, 02:42:28 pm »
    I'd guess they probably lean to whatever they feel is conservative. . .if the area needs rain, they'll say 2-4 inches even if they think it might be 5. . . .though in the case of rain it's still pretty important to warn of possible flooding and such.

    We could find this out by watching our local forecasts AND checking the NWS and figuring out which way they go. :D
    My Kindles
    Hermoine's Handbag (Voyage)
    Ed's (Voyage Refurb)
    Bedtime Reading (Oasis 9)
    Godric's Hollow (Basic 7)
    My Tardis (PW 10)

    Offline Eltanin Publishing

    • Status: Scheherazade
    • *****
    • Posts: 1266
    • Charlotte, VT
    • Cathy Ryan - Editor-in-Chief
      • View Profile
      • Eltanin Publishing
    Re: Week 3: Chapters 4 & 5
    « Reply #4 on: November 29, 2012, 06:07:04 am »
    I'd guess they probably lean to whatever they feel is conservative. . .if the area needs rain, they'll say 2-4 inches even if they think it might be 5. . . .though in the case of rain it's still pretty important to warn of possible flooding and such.

    We could find this out by watching our local forecasts AND checking the NWS and figuring out which way they go. :D

    The thing is, our local weatherman (at least, my favorite of the couple locals) never gives a probability of precipitation (a percentage). The graphic just shows pictures and highs and lows, and then verbally he'll talk about a chance of rain or snow (using words like "slight chance" or more decidedly "it's going to rain late this evening..."). Other than forecasted highs and lows, the only other firm numbers I can think of that he gives is when there is snow predicted and he shows a map with colored blobs indicating how many inches for different areas. That would certainly be testable.

    I "followed" our local weatherman on twitter yesterday and sent him the message "@TomMessner have you read Nate Silver's The Signal & the Noise, & his chapter on weather forecasting? Do your forecasts tend to lean "wet"?" and he replied, "I have not. But, now I'm curious."

    Edited to add: We're supposed to have just a little snow tonight, so I compared the NWS page with my local tv station's forecast. NWS says this evening, less that a half inch, overnight - less than an inch. So a total of less than 1.5 inches. Local station shows a map, and we're located right between where they've written "Dusting to 1 inch" and "1 to 2 inches". So they seem to be pretty much in agreement.
    « Last Edit: November 29, 2012, 06:20:21 am by Eltanin Publishing »
     

    Offline drenfrow

    • Status: Scheherazade
    • *****
    • Posts: 1217
    • Gender: Female
    • Terlingua, Texas
      • View Profile
    Re: Week 3: Chapters 4 & 5
    « Reply #5 on: November 29, 2012, 01:26:47 pm »
    One thing I took from the weather chapter was realizing how pointless it is to check the 10 day forecast as I do every morning online.  This week I have been making myself look at only the next few days knowing that much farther than that is just too uncertain.
      


     
    Woke up this morning with a terrific urge to lie in bed all day and read. -Raymond Carver

    Offline Eltanin Publishing

    • Status: Scheherazade
    • *****
    • Posts: 1266
    • Charlotte, VT
    • Cathy Ryan - Editor-in-Chief
      • View Profile
      • Eltanin Publishing
    Re: Week 3: Chapters 4 & 5
    « Reply #6 on: November 30, 2012, 07:11:28 am »
    Last night my husband was annoyed because the predicted snow showers turned to rain for an hour or two. He said, "NONE of the websites said ANYTHING about this snow changing to rain!" So, he was upset that they had essentially given a 0% chance of the precipitation being rain (since none of them mentioned that option) and yet that's what happened. I told him that, you know, when it's close to freezing, you just never know, and yet he's right that it's odd that none of the forecasts mentioned this possibility. It made for very icy roads as it all froze up.

    And yet, Silver keeps saying how weather forecasting is one of the few success stories in the book, where it really has improved over the years. Perhaps our expectations of accuracy have increased as the accuracy has.
     

    Online Ann in Arlington

    • Global Moderator
    • Status: Shakespeare
    • *****
    • Posts: 67017
    • Not really a moderator. :)
      • View Profile
    Re: Week 3: Chapters 4 & 5
    « Reply #7 on: November 30, 2012, 10:37:27 am »

    And yet, Silver keeps saying how weather forecasting is one of the few success stories in the book, where it really has improved over the years. Perhaps our expectations of accuracy have increased as the accuracy has.

    I think you're right about that. . .hey, did you check the NWS to see if they said it might be partly rain? ;)

    I think people see those 10 day forecasts and think they're really accurate. . .what they don't realize is that, if they look at them 10 days in a row, the forecast changes slightly each day.  So they get to the end of the 10th day and think, wow, it happened just like they said it would a week ago. . .but, in fact, it's not the same forecast today for tomorrow as it was a week and a half ago for tomorrow, but the change all week was subtle and mostly not paid close attention to.
    My Kindles
    Hermoine's Handbag (Voyage)
    Ed's (Voyage Refurb)
    Bedtime Reading (Oasis 9)
    Godric's Hollow (Basic 7)
    My Tardis (PW 10)

    Offline Eltanin Publishing

    • Status: Scheherazade
    • *****
    • Posts: 1266
    • Charlotte, VT
    • Cathy Ryan - Editor-in-Chief
      • View Profile
      • Eltanin Publishing
    Re: Week 3: Chapters 4 & 5
    « Reply #8 on: November 30, 2012, 01:14:12 pm »
    I think you're right about that. . .hey, did you check the NWS to see if they said it might be partly rain? ;)


    NWS (http://www.weather.gov) is the main place I check for weather online. I also love their "Hourly weather graph" (link on the bottom right after you put in your zipcode for a forecast). I am often wondering, "yes, it's likely to rain today, but WHEN?" Also, for total weather nerds, there's a "Forecast Discussion" (link at the bottom of the page), which can get a bit indecipherable with all the acronyms, but it can give a sense of how the human forecasters are interpreting stuff.

    And nope, no mention of the precip being rain (at least, on the main page - I didn't look at the graph yesterday).
     

    KBoards.com

    • Advertisement
    • ***