TOBS For Dummies

Imagine you are operating a temperature station in July, 1936 – the hottest month in US history. Your time of observation is 5 pm. The temperature is 109 degrees, and you reset your thermometer.

At 6 pm, a cold front comes through and the temperature drops to 60 degrees.  The next day is 40 degrees cooler, but your min/max thermometer reads a maximum of 109 degrees, from when you reset it a day earlier. Are you actually going to enter 109 degrees for the high that day?

Of course not. Any responsible person will reset their thermometer at night, regardless of the time of observation. TOBS is not a subtle problem. It is a very much in your face problem as the station operator, and I have a very hard time believing that many people would be irresponsible enough to reset their thermometer at 5:00PM, and then not reset it again until the next afternoon. It doesn’t pass the sniff test of human psychology.

About stevengoddard

Just having fun
This entry was posted in Uncategorized. Bookmark the permalink.

41 Responses to TOBS For Dummies

  1. darrylb says:

    Of Course—Just a little story of interest–In what was known as the Armistice Day, now Veteran’s day blizzard, Southern MN.
    People were 15 minutes from home, 70 deg F., calm, And when the storm hit, many could not make it home, they simply froze to death. Dropped down 50 degrees, with gale like minutes.
    Weather happens.
    A great shortcoming of how history is taught is the omission why and how events occurred as precipitates of weather (pun intended) A great shortcoming of climate studies is the omission of the accurate history of weather. and climate.

  2. nickreality65 says:

    How accurate is chaos?

  3. Shazaam says:

    It would seem that Toto has not had the pleasure of dealing with many County or State government employees.

    I can picture a government hack behaving in that manner. They reset the min-max because they have to. Not because they give a flying flip about the accuracy of the data they record. Thus, such resets will happen at their convenience.

    I do agree that no rational individual who cared about the data would behave in such a matter. And then we have the others…….

    • Ernest Bush says:

      You live in an era where irrational and immoral behavior is celebrated by both government and private economy employees.. Do not judge those of the past by applying today’s standards. My father-in-law was an honorable man who worked for the IRS. He died in 1991 at age 72. He would be humiliated and outraged by the behavior of the IRS and so many of its employees today. Many before him would be outraged at what was being done to their careful temperature readings today.

      • Brian G Valentine says:

        There aren’t so many, the civil servants of the Federal Government are honorable as far as the many i have known, there are a very few political appointees who make decisions that are less than well considered.

      • Shazaam says:

        I work in a township openly famous for purchasing 2 or 3 times the gravel that gets used annually on the roads.

        I also served a term on a grand jury where, when we asked about inditing one of the many cops who blatantly lied under oath we were told that the prosecutors would likely be “to busy” to pursue the matter. (In this state, if no action is taken by the prosecutors in 270 days, the indictment is dropped and expunged)

        Thus I may have a rather jaded view of “government integrity”.

        I know an IRS employee who is as disgusted with what is happening. This the only IRS employee I’ve ever heard of who follows the “no-harm-no-foul” principle. Many of those audited have been told how to fix their minor mistakes without a punitive penalty. The willful frauds receive a different treatment. Vanishingly rare behavior in that organization, and gone with yet another retirement someday.

  4. Bob Greene says:

    When and if the thermometer is reset depends on who, and how many, take the readings. For example, I read that some weekend readings were missed because the reader wasn’t there. I’d bet it is more likely that the readings and resets are done about the same time of day. A sudden drop in temperature is not likely to get caught. And I’d even bet the general thought would be that all those errors would average out.

    It’s very unlikely that most of these manual read stations have a data collector whose primary job is collecting that data and it won’t always be done very precisely. The collections will likely be fairly good, but the precision and consistency won’t be like you’d find in a QC lab.

    • Gail Combs says:

      Most of the stations were run by volunteers not government hacks so I would expect more dedication to the job due to a passion for science.

      …Today weather observers who make up the Cooperative Observing Program are a core of volunteers numbering over 10,600. There history extends back long before there was a National Weather Service.… This network of volunteer weather watchers has become an American institution. The program and its history of success continue to arouse the envy of other countries around the globe. It has been acclaimed as the most cost-effective weather data collection network in the world.

      The all volunteer weather observer network was envisioned by Thomas Jefferson in 1776 when he began to recruit observers in Virginia. By 1800, his vision has spread to five other states. By 1891, the network had grown to nearly 2,000. In 1856 the Honorable Isaac Newton, U.S. Commissioner of Agriculture, endorsed a recommendation by the Secretary of the Smithsonian Institution, Prof. Joseph Henry, that a more extensive weather service be established for the benefit of agriculture. Five years later the U.S. Weather Bureau became a new agency under the U.S. Signal Service. On February 2, 1870, President Ulysses S. Grant signed a joint resolution of Congress authorizing the establishment of a national weather bureau. With the Organic Act of October 1, 1890, all weather functions were transferred to a new agency called the Weather Bureau. Among other duties, one of the primary mandates of the Organic Act was the volunteer weather observer program. This program has evolved into the National Weather Service (NWS) Cooperative Weather Observer Network. Although the network has been very successful in fulfilling its original agriculturally oriented mission of defining the weather and climate of the United States, its data are now used for a myriad of things including water and land management, recreation, environmental impact studies, litigation and insurance, energy production and energy use, engineering, architectural design and construction, and agriculture and farm management, to mention a few. Not only are the Cooperative Weather data invaluable to the NWS, but these data are also the basis for critical business decisions in the private sector. Without these data, the National Weather Service would find it nearly impossible to fulfil its mission…..

  5. QQ says:

    We don’t know what happened in June, 1936… unless we want to dismiss TOBS.

    Then we know that every coop station was located within convenient walking distance of every volunteer observer’s home. Not at the ass end of their farm, not at their workplace, not at an airport or train station, just right outside their back door. We know that every observer, in addition to the observation time recorded in the metadata, took a second, unrecorded trip out to the station to reset the thermometers after the peak temperature hour passed. We know this happened without fail, even for the subtle days, and not just the 40-degree reductio ad absurdum drops. And we know this happened despite several contemporary papers from the 1900s through the 1950s noting the problem of TOBS (and some even making primitive attempts at bias correction).

    There’s definitely some human psychology that doesn’t pass the sniff test here, but it’s not what you think it is. And that’s not even touching on the massive TOBS/UHI doublethink at work.

    • Those 12,000 people who died from the heat during one week in July, 1936 simply didn’t understand that the heat was double counted, and it really wasn’t hot.

      Thanks for clearing that up

      • Brian G Valentine says:

        Quadrupling electric bills for A/C ought to fix that problem up in no time.

      • Bill H. says:

        Nice sarcastic rant, Tony. Now perhaps you could address the substance of QQ’s comment.

        • Chip Bennett says:

          Substance? What substance?

          We don’t know what happened in June, 1936…

          (Unless we want to make arbitrary adjustments to raw data)

          If we don’t know what happened, we don’t know what happened. In my world, we have a very simple rule: if it wasn’t documented, it didn’t happen. So, even if you theorize that TOBS exists for certain raw data from 80 years ago, unless you have documented evidence of that error, you can’t use it as a basis for manipulating the data.

    • _Jim says:

      Maybe QQ can explain why there are continued changes to past data (not the raw data but the adjusted data) WHEN all this was ‘set in stone’ because “ several contemporary papers from the 1900s through the 1950s noting the problem of TOBS ” were written and one should thinking CONTINUED adjustment of the adjusted data would not be needed.

      Adjust once, and that would be it. But, no. Continued adjustment to this day, of the adjusted data.

      .Why are they doing that QQ?


      • QQ says:

        None of the various attempts at TOBS corrections were officially incorporated into the record until USHCN v1, when the method of Karl 1986 was adopted. Up until that point all of the methods that had been explored were … exploratory. They didn’t go beyond the papers they were tried in.

        Of course, this was all clearly noted in Hansen 2001 and the USHCN v1 documentation, which lay out the rationales for That One GIF, but I’m guessing we’re of the opinion that anything short of writing in hot pink letters across the trendline of the chart: “ADJUSTED FOR TOBS AND MMTS AND UHI (BUT NOT ENOUGH UHI [CONSULT STEVE GODDARD’S CRYSTAL BALL FOR EXACT COEFFICIENTS])” is literally Lysenkoism.

        • The National Weather Service has measured more than six degrees UHI in Phoenix, so clearly 0.1 degrees has that covered. Your comment is literally Goebbelism

        • QQ says:

          Whoa, I see we’ve graduated from refuting my comments on TOBS with single stations to refuting them with whole cities! They grow up so fast 🙂

        • _Jim says:

          Question not answered; why are incremental adjustments performed each month?

        • Studies show three degrees in New York and three degrees in Barrow, AK . Weird how UHI only targets cities where they actually take measurements of it.

        • _Jim says:

          QQ, if you don’t know why they continue to ‘cool the past’ with each successive adjusted data release, just say so. Please don’t just hand wave …

        • QQ says:

          In New York?! My God, the UHI effect on that trend must be enormous! New York was a pristine wilderness in the 1930s! And Barrow?! Not Barrow! I shudder to think of the effects such a UHI must have on the CONTIGUOUS UNITED STATES temperature record.

          Do you actually intend on collating any of this smoking gun data into some sort of mathematically verifiable conclusion, or are you just going to pull low-effort cherry-picks from now unto infinity to go with your thrice-weekly rotation of the same images since 2012?

        • In most fields of science people do sanity checking calculations to see if their assumptions make sense – before they start coding. 0.1F UHI is completely out of the ballpark of any rational number.

          I saw 5-10 degrees on my bicycle going from open space into the gym parking lot tonight. That is fairly typical.

        • QQ says:

          And if that parking lot and every other morsel of urbanization in your city had sprung up overnight, maybe you’d have a point. But now you’re comparing an absolute difference you noticed on your bike to a difference in an annual mean temperature trend. What would that five-degree difference in temperature you notice look like as a trend? Let’s say across, oh, fifty years of urbanization and development? I’ll leave the math to you, the expert.

        • Every night the weatherman says “and a few degrees cooler in outlying areas.” He must mean to say “0.1 degrees cooler” because that is the official NCDC UHI adjustment.

        • RealOldOne2 says:

          “Summer land surface temperatures of cities in the Northeast were an average of 7°C to 9°C (13°F to 16°F) warmer than surrounding rural areas over a three year period, the new research shows. The complex phenomenon that drives up temperatures is called the urban heat island effect. Heat islands are not a newly-discovered phenomenon. Indeed, using simple mercury thermometers, weather watchers have noticed for some two centuries that cities tend to be warmer than surrounding rural areas. … The compact city of Providence, R.I., for example has surface temperatures that are about 12.2°C (21.9°F) warmer than the surrounding countryside.” – NASA release ( )

          QQ, Can you please show us where any of the land-ocean temperature records for Providence, R.I. have made a correction for this double digit UHI effect which has been documented by the NASA satellites? That would be either warming the temps a century ago or cooling the present by several °C.
          While you’re at it, we’d appreciate showing the multiple °C UHI adjustments for all those other cities in the Northeast.

          Thanks QQ. I’ll be waiting on the edge of my seat. (NOT!)

        • RealOldOne2 says:

          Whoops. Missed the /i at the end of the NASA quote

        • QQ says:

          Just as soon as you explain to me how an effect “noticed for some two centuries” per your own damn citation suddenly started impacting Providence between the 1930s and the 2000s. Was urbanization in the Northeast not a thing until the mid-20th century? Judging by the amount of time you people spend talking about ABSOLUTE urban/rural differences rather than TRENDS, that seems to be the official position here.

          That’s why the UHI TREND effect is small, and why stacking up a jillion papers saying “in City X it’s 15 degrees hotter than the countryside” just makes you look like idiots. Well that and every other thing you post.

        • Brian G Valentine says:

          “Urbanization” of a city such as Providence by such means as “asphalt” of extensive areas was not a popular idea before 1930, indeed.

          Why not pick another blog to demonstrate your genius? You aren’t doing an especially good job of it here

        • RealOldOne2 says:

          QQ: “your own damn citation … suddenly started affecting..”
          Now, now, temper, temper. Did you miss your anger management class?

          Knowing about an effect and understanding the magnitude of the effect are two different things. I never said it suddenly started affecting, just that it has certainly impacted the temperatures over the last century, probably on the order of up to a few °C.

          California state climatologist James Goodridge stated: “In California the rate of increase in temperature commonly attributed to greenhouse warming was 3.14°F century-1 for 29 stations located in counties with populations over 1 million people and 0.04°F for 27 stations located in counties with fewer than 0.1 million people. Long term temperature trends are clearly a function of urban population density.” – Goodridge 1996, BAMS

          And significant UHI effect has been documented in towns of <5,000 people.
          "The village of Barrow, Alaska, is the northernmost settlement in the USA … The population has grown from about 300 in 1900 to more than 4600 in 2000. … Data loggers (54) were installed in the ~150 km2 study area to monitor air and soil temperature … During the winter (December 2001 – March 2002), the urban area averaged 2.2°C warmer than the hinterland. The strength of the UHI increased as the wind velocity decreased, reaching an average value of 3.2°C under calm (<2 m s-1) conditions and maximum single day magnitude of 6°C." – Hinkel 2003, Intl Jrnl of Climatology

          What is very typical over the last century is that urban growth has happened around temperature stations causing an unadjusted-for UHI, then the station is moved to a "cooler" location near the edge of the town, then over a few decades, urban growth happens around that area causing additional UHI, & the process repeats. Instead of properly adjusting for UHI, the opposite is done, as the previous station's data is cooled to match the overlap temperature of the new location. This cumulative effect is to falsely amplify the UHI by causing false warming trend where one didn't exist or was of less magnitude.

          Bottom line is that the present adjustments of cooling the past and warming the future is exactly the opposite of what should be happening to properly adjust for UHI in the temperature records of the past century.

          Your denial of reality by sticking your head where the sun doesn't shine, pretending that UHI isn't a factor or that it has been adjusted for or that urban growth hasn't happened, is typical doomsday cult zealot behavior when your cherished cult religion is presented with empirical science which you can't bring yourself to accept because of your ideological commitment to your cult dogmas.

          So rather than raise strawmen & red herrings and deny the reality that is presented to you, cite examples where UHI adjustments are made in the proper magnitude and direction. I’ll be waiting.

    • Ernest Bush says:

      This is precisely where the anecdotal evidence contained in newspapers and magazines of past eras come into use. We know that the little ice age really happened because of engravings and personal accounts. We know that the Vikings lived and died while farming in Greenland due to archeological evidence. Tree roots were found to have grown through the bodies of buried Vikings, so we know it was warm and wet enough to sustain them. Proxy data from that far back merely confirms what was already known.

      You can criticize the readings from thermometers of the time, but thousands of newspaper accounts exist of the effects of the 1930’s extreme heat. Some idiots would accuse Steve of cherry picking with the above newspaper article, except that he has presented similar reports from newspapers around the country and the world repeatedly since I have been reading his blog.

      Perspective has been lost in the past decades, also. There are bitter arguments raging over tenths of a degree changes over the pasta 150 years. I have used an old, large, bulb thermometer since 1971 made by Kodak to measure liquid temperatures. It is more accurate than any mechanical dial thermometer I have ever used. I can estimate using it to an accuracy of quarters of a degree and it compares well for accuracy against modern thermometers with digital readouts. Having precise digital readouts has nothing to do with the accuracy of an instrument.

    • Chip Bennett says:

      The correct way to handle theoretical error in the raw data is to establish a margin of error around the raw data, not to change the raw data arbitrarily.

      The only way to “change” raw data is to have identical, corroborating measurements that prove the magnitude of the error – in which case you’re not actually changing anything; you’re just justifying the use of alternate raw data.

  6. Send Al to the Pole says:

    Just as importantly, WHO could possibly know what they did, such that they would make the assumptions that NOAA/GISS are making? They dreamed up a convenient story and called it science.

  7. Larry Fields says:

    Thanks for following up on my ‘for dummies’ request.

  8. Brian G Valentine says:

    One reason i put faith in satellites only is the very obvious correction for humidity. Daily averages are arithmetic means of daytime highs and night time lows, if it is humid, the night temperature is going to be higher, and there is no uniform way to adjust for that.

    (Golf courses in California deserts that are constantly watered has a lot to do with the appearance of California increasing “warmth”)

    • Ernest Bush says:

      Due to an extreme lack of humidity, evaporation of ground water actually cools the southwestern deserts at night in watered areas until the monsoon season brings in humidities of 40 to 60 percent.

      I live in Yuma, Arizona, in a place called the Valley. The water table is high and there is evaporation of water from all the lush grass and trees grown around here. You have to be up high to notice the sand dunes beyond the city. Often when I was driving toward my job at 4:30 am I had the window rolled down to enjoy the morning cool. Then 16th St. went up a hill onto the so-called Mesa and suddenly it was quite warm and time for air conditioning. The difference was dramatic. You can climb out of an 88-degree swimming pool when it is 110 out and the dry wind will cool you until you start shivering. The humidity will be in single digits.

      It is actually cooler to drive through irrigated fields than through desert terrain or the city. The humidity from the monsoon season raises the low temperature at nights because it traps the heat. There is no cooling from a phase change to be had. I would add that monsoonal humidity increases usually cool the high temperature by a few degrees. It does nothing for the misery by dropping it from 115 to 112, however, with 60 percent humidity.

  9. JP says:

    Say, I’m a weather observer at the Dtichweed Airport in Ditchweed Indiana. It is 1936. I divide my time at the air port between sweeping out base operations and taking the daily weather reports. Or is it hourly? That is the problem. I spent 12 years in the Air Force in the weather field. Two of those years were as an observer (because of the shortage of trained forecasters in the very early 1980s, I was selected to get training as a forecaster after only 2 years in the field). In those days, we didn’t earn much money (about $4500/year). One thing I noticed immediately in the weather field was the difference between various weather reporting stations. Believe it or not, most of the reporting stations had 24×7 operations. Until the advent of more advanced weather equipment such as the TMQ-11, weather observers took a wet-bulb and dry bulb temperature every hour (by FAA rules they could take their measurements up to 10 minutes early, as long as they transmitted their report on the hour). The only thing the observer had to really know how to do was calculate the dew point temperature.

    The growth of reporting stations really came about through the spread of aviation. And the old Weather Bureau used this expansion to also expand its data sets. In the old days, one could be an FAA weather observer ones entire career. There were no shortage of former Navy and Army weather observers trained to take official FAA weather reports – and they reported at the very least EVERY HOUR. So, in 1932-1937 there were literally hundreds of air ports in the US that reported hourly weather info (cloud cover, visibility, temp, dewpoint, altimeter, sea level pressure, etc… There would be no need to apply a TOB adjustment, as the Weather Service had an ongoing 24 hour, 365 day data set for each of these air ports. Granted, the number of reporting stations that had humans taking hourly dry bulb/wet bulb temps began to fall off the map in the 1980s. Auto-observing stations, and civilian MMT reporting stations began to fill the maps. But, that doesn’t mean we have to be at the mercy of the ridiculous Tmax and Tmins. We have the technology to allow for hourly MMT and COOP temps.

    Why in the world NOAA decided to only record 2 daily temps (Tmax., Tmin)? It should record hourly Tmax/Tmins for a period beginning at 0000Zulu to 2359Zulu. It’s computers can then take the highest hourly Tmax and lowest hourly Tmins and use them as the official Max/Min for that 24 hour period. Ergo, there would be no need for TOBs, anywhere.

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s