Thanks for directing me to the CA thread. I found a copy of Karl, et al, 1986 in Hu McCulloch’s comment #110:

http://ams.allenpress.com/archive/1520-0450/25/2/pdf/i1520-0450-25-2-145.pdf

I read the article hurriedly but I need to give it more thought. My first impression is that TOBS may introduce considerable distortion to the areal data. For example, the month-to-month variability for Mohonk Lake listed in my 10/19 message, especially in the 1956-2006 interval, doesn’t seem logical to me. This variability may be due to the spatial patterns of δ shown in Fig. 6 of the article. The gradients for δ are quite steep during DJF in the Mohonk Lake region. The iso-contours also shift quite rapidly.

I found nothing in the TOBS article that explains the significant step-wise changes in the difference between the areal and COBS data occurred in 1909 and 1955 for Mohonk Lake. I checked several monthly B91 Forms before and after 1909 and 1955 and there were no significant change in Time of Observation. (In 1954 and 1956, the time listed on the B91 forms was 4:00 PM; in 1908 and 1910, the time listed was either 6:00 PM or “about sunset.”)

In the Fort Collins example that I mentioned where there was no difference between the areal and TOBS data between 1905 and 1940 … I guess the TOBS “adjuster” must have been “OFF” during that interval!!

Needless to say, I am still puzzled by the inexplicable step-wise changes that I have observed. I’ll examine some more sites when I have time and see if a similar pattern persists.

BTW, I noticed that many of the comments on the CA thread referred to COBS as though it was used to adjust the daily TMAX and TMIN data. Since the B91 Forms list only the time of observation for the month, it seems clear to me that COBS is a monthly adjustment.

Thanks, also, for the guidance on how to export a graphic from Excel to WordPress. I’ll try using it to display some of the step-wise changes that have been puzzling me.

Is COBS a typo, and you meant TOBS?

I cannot explain the algorithm behind the TOBS adjustment, because I have not obtained a copy of Karl, et al. 1986; Vose et al., 2003 (see the links in the main body of this post). TOBS has been discussed at length on Climate Audit, particularly on this thread. You should be able to pick up some interesting insight from the comments (I myself need to reread it). In particular, one of the early comments points you to a pretty good description of TOBS here.

As for getting an Excel chart into WordPress. Here is a very simple procedure:

1. Click on the chart object, then press CTL-C to copy the chart (make sure you click on the chart near the edge, and not a location closer to the center, otherwise you end up with just the plot area or an axis.)
2. Open Microsoft Paint. An empty artboard will appear.
3. Press CTL-V to past the chart image into paint.
4. From Paint, click File>Save As and save as a GIF image into a directory of your choice.

Now, when you go into WordPress you can upload your GIF image and reference it.

I should have indicated that the first part of my message pertained to Mohonk Lake.

TOBS continues to be a mystery to me. I have compared the areal data with the COBS data in hcn_doe_mean_data file for the 1900:2006 interval and was perplexed by the results. Significant step-wise changes in the difference between the areal and COBS data occurred in 1909 and 1955. All difference values were constant between the changes as listed below:

MONTH 1900-08 1909-55 1956-2006
JAN 1.10°F 1.10°F 1.40°F
FEB 1.10°F 1.50°F 1.80°F
MAR 0.70°F 0.90°F 1.30°F
APR 0.60°F 1.10°F 1.50°F
MAY 0.50°F 1.30°F 1.60°F
JUN 0.30°F 0.88°F 1.20°F
JUL 0.30°F 0.80°F 1.00°F
AUG 0.50°F 1.00°F 1.20°F
SEP 0.70°F 1.31°F 1.60°F
OCT 0.80°F 1.11°F 1.50°F
NOV 1.30°F 1.31°F 1.70°F
DEC 0.90°F 0.90°F 0.90°F
AVE 0.72°F 1.10°F 1.41°F
DJF 1.00°F 1.18°F 1.43°F
MAM 0.58°F 1.09°F 1.47°F
JJA 0.37°F 0.90°F 1.13°F
SON 0.94°F 1.23°F 1.60°F

I also compared the areal data with the COBS data in hcn_doe_max_data and hcn_doe_min_data files for the 1900:2006 interval and observed the same step-wise pattern, although the values were different.

I have also made the same comparisons for other surface station and observed similar perplexing differences between areal and TOBS data. For example, Fort Collins, CO (053003) for the 1900:2006 interval displayed significant step-wise changes in TAVG, TMAX and TMIN in 1905 and 1940 but there was no difference between the areal and TOBS data between 1905 and 1940.

Can anyone explain to me how NCDC calculates TOBS?

REPLY: Yes and no, you can get the absolute RAW B91 data forms free, but you have to hand transcribe. If you want a CD of data, then yes you have to pay for it to NCDC. – Anthony

http://www.guardian.co.uk/flash/0,,1656541,00.html
Clean water ,clean energy ,and the solution for the Greenland gulfstream slowdown:

Put in sea a construction of windmills combined with electric boilers, these get preheated by suncollectors and sunmirrors.
the steam they produce gets used by steam engines wich produce again electricity ,salt and when you condesate the steam coming from the steam engines sweet water.
You take out salt water coming from the warm gulfstream and the left over salt from the steam engines gets used to release in the cold stream.
So the cold stream gets saltier again and and won’t mingel with the warm gulfstream becaurse before it was less saltier caursed by the melting gletchers and slowing down the warm gulfstream becaurse less saltier water is lighter.
Preventing an possibel iceage.
The sweet water can then be used to make hydrogen ,so if there is for exampel a lot of wind and you don’t need all that electricity , you are abel to stock it and use it later.
The rest of the steam you blow in the air.
The oxygen deriving from turning water into hydrogen can be devided in sea to clean the water and to make more suitabel for marinelife.
So you produce electricity and /or hydrogen

I should have noted that under “…some of my observations:” the step-wise changes in Adj RAW TAVE, Adj RAW TMAX and Adj RAW TMIN were with respect to the average annual temperature.

Also, another correction:

3. Adj RAW TMED = Adj RAW TMAX – Adj HCN TMAX

should have read

3. Adj RAW TMED = [Adj RAW TMAX + Adj RAW TMIN]/2

Again, sorry about the confusion. I guess I was in too much of a hurry to go to dinner.

HCN del TAVE = HCN TAVE –HCN TAV3

should have read

HCN del TAVE = HCN TAVE –HCN MED

Sorry about that …

Thanks again for your insights. I have commented previously on this blog that I have been puzzled by what appeared to me to be step-wise adjustments that NCDC makes to surface station data. As I’m sure you know, WRDC archives surface station data which I compared to USHCN data for several surface stations and observed the step-wise differences. I wondered if the similar step-wise differences might be observed between NCDC’s daily and monthly data sets for Mohonk Lake,NY (305426). I have completed my analysis and observed similar step-wise differences between the daily and monthly data sets.

I use Excel but, unfortunately, I haven’t learned how to export Excel graphics to WordPress. It would be much easier to convey what I have observed if I could do so. If you permit me, I’ll try to briefly explain what I have observed.

First, I refer to the HCN data set as the RAW data set. I updated the RAW data set by correcting the -77°F transcription error and using the B91 Forms to fill the data cells for the missing six days. While doing that, I realized that the entire B91 Form for April 1965 had not been transcribed properly. Therefore, I incorporated the B91 Form data.

I then calculated the monthly averages for TAVE(F), TMAX(F) and TMIN(F), for the interval 05/1948 through 12/2005 and compared the results with the corresponding monthly values that I downloaded from
http://cdiac.ornl.gov/cgi-bin/broker?_PROGRAM=prog.climsite_monthly.sas&_SERVICE=default&id=305426

I defined the following adjustments:

1. Adj RAW TMAX = RAW TMAX – HCN TMAX
2. Adj RAW TMIN = RAW TMIN – HCN TMIN
3. Adj RAW TMED = Adj RAW TMAX – Adj HCN TMAX
4. Adj RAW TAVE = Adj RAW TMED + RAW del TAVE – HCN del TAVE

Where

1. RAW del TAVE = RAW TAVE – RAW TMED
2. HCN del TAVE = HCN TAVE –HCN TAV3

Briefly, here are some of my observations:

1. Stepwise adjustments occurred in 1954-56 and in 1990.
2. Steps-wise changes in Adj RAW TAVE:

Average 1949:1954 = 0.39°F
Average 1956:1980 = 0.63°F
Average 1991:2005 = 0.03°F

3. Steps-wise changes in Adj RAW TMAX:

Average 1949:1954 = 1.27°F
Average 1956:1980 = 1.85°F
Average 1991:2005 = 1.20°F

4. Steps-wise changes in Adj RAW TMIN:

Average 1949:1954 = 0.92°F
Average 1956:1980 = 1.02°F
Average 1991:2005 = 1.62°F

So … does this make any difference from a climate change perspective? I don’t know. For the 1949:2005 interval, simple linear regression shows the following:

RAW TAVE = 1.79°F
HCN TAVE = 1.55°F
RAW TMAX = 2.56°F
HCN TMAX = 2.97°F
RAW TMIN = 1.01°F
HCN TMIN = 0.19°F

I guess that it depends upon on your perspective.

The second sentence in the last paragraph should have read: “It turns out that the value of 77°F recorded on the B91 Form for second day of that month…”

Thanks, I’m a novice in this field and do not fully understand NCDC’s homogenization, TOBS, and FILNET adjustments. Here are a few of my comments and observations:

Homogenization: I really don’t know anything about that aspect.

TOBS: TOBS is a mystery to me. I read BarryW’s September 24th, 2008 at 12:50 pm comment to your post on CA and I have read the general discussion of the subject on the USHCN website. Referencing al station data to midnight may be logical but it is not apparent to me how that adjustment is made.

FILNET: Surprisingly, the 1948:2005 daily RAW data set with 3X21062 data points has only 14 missing data points; two days where TAVE and TMAX are both missing, two days where TAVE and TMIN are both missing and two days where TAVE, TMAX and TMIN are all missing. On each of these six days, the values for TMAX and TMIN were listed on the B91 Forms. Therefore, it doesn’t appear to me that FILNET was of much use.

As regards Mohonk Lake, I examined the 12 B91 Forms for 1951 and noted that 4 PM was the Hour of Observation on all forms. For 1951, the monthly differences between RAW and HCN temperatures (∆Ts) are:

MONTH ∆TAVE ∆TMAX ∆TMIN
JAN 1.33°F 1.03°F 1.08°F
FEB 1.64°F 1.56°F 1.30°F
MAR 1.19°F 1.01°F 0.85°F
APR 1.32°F 1.23°F 0.88°F
MAY -1.05°F -3.21°F 0.73°F
JUN 1.05°F 1.09°F 0.69°F
JUL 1.04°F 1.16°F 0.54°F
AUG 1.25°F 1.32°F 0.76°F
SEP 1.53°F 1.66°F 0.93°F
OCT 1.39°F 1.13°F 1.18°F
NOV 1.42°F 1.32°F 1.22°F
DEC 1.06°F 0.88°F 0.91°F
AVE 1.10°F 0.85°F 0.92°F

I chose 1951 to review because of the negative values for ∆TAVE and ∆TMAX in May. It turns out that the value of 77°F recorded on the B91 Form for that month was transcribed in the RAW data set as -77°F. It appears to have been corrected by NCDC before including it in the HCN data set.The reasons for the other differences are not apparent to me.

Reply by John Goetz: The homogenization algorithm does catch the type of outlier you describe (the transcription error). However (and I have not seen the code), I do not believe it will necessarily take a -77°F value and change it to 77°F. It might change it to a 63 or 84 or some other number that is based on the values of the “closely correlated” nearby stations.

Ideally, NOAA would spit out a list of the outliers and have a transcriptionist go back and manually double-check the B91. They just need to do that once, and then once a month after that on all newly-added data.

Here comes a stuuupid question, To Goetz or anyone:
Goetz you write:

“Here is a plot of the difference between the monthly raw data from Mohonk Lake and the data GISS creates in GISTEMP STEP0 (yes, I am well aware that in this case it appears the GISS process slightly cools the record). ”

Hmmm…

If GISS in 1880 is 0,4 degrees COLDER than Mohonk
- and in 2000 is 0,0 degrees COLDER than Mohonk

Yes, then the COOLING done by GISS is bigger in 1880 than 2000 ?

And then GISS has induced a WARMING TREND !?

- ups i have the feeling that i got something wrong, but i have to mention…

K.R. Frank

Reply by John Goetz: By cooling the record, I mean the trend from 1896 – 2006 in GISS is slightly less warm than the trend in the raw data. GISS actually warms the older temperatures and cools the later ones, thus reducing the overall trend.

I tried to post a comment a short time ago but I didn’t receive the usual response “awaiting moderation.” I tried subnitting it again but still no response.

Did you receive anything from me?

hmccard

Reply: No – Anne

More: Nevermind, found them in the spam bucket – Anne

I’m sure that you know that the daily temperature data from the Mohonk Lake surface station can be found for the interval 05/1948 through 12/2005 at

http://cdiac.ornl.gov/cgi-bin/broker?_PROGRAM=prog.climsite.sas&_SERVICE=default&id=305426

I selected a random sample of ten completed B91 Forms for Mohonk Lake from the site that you referenced

http://www7.ncdc.noaa.gov/IPS/coop/coop.html?_page=2&state=NY&foreign=false&selectedCoopId=305426&_target3=Next+%3E

and verified that the values for TMAX(F) and TMIN(F), with one exception, were identical. The exception which occurred on 07/12/1956 is an obvious transcription error; the value (67°F) listed in the “At OBSN” column was recorded for TMIN(F) instead of the correct value (64°F). In addition to TMAX(F) and TMIN(F), the USHCN site also lists TAVE(F) which equals [TMAX(F) + TMIN(F)]/2 rounded to the next higher integer when either TMAX(F) or TMIN(F) is an odd integer. Therefore, I believe the data archived at the site that I referenced is the “RAW” daily temperature data for station 305426.

So far, so good …

I then calculated the monthly averages for TAVE(F), TMAX(F) and TMIN(F), for the interval 05/1948 through 12/1959 and compared the results with the corresponding monthly values that I downloaded from

http://cdiac.ornl.gov/cgi-bin/broker?_PROGRAM=prog.climsite_monthly.sas&_SERVICE=default&id=305426

How did they compare? Not so good!!

The average differences for the 139 month interval are (s.d. shown in parentheses):

RAW TAVE(F) – HCN TAVE(F) = 1.39°F (0.43°F)
RAW TMAX(F) – HCN TMAX(F) = 1.43°F (0.62°F)
RAW TMIN(F) – HCN TMIN(F) = 0.83°F (1.17°F)

What adjustments do you think were made by NCDC that caused these differences?

I may add to this post after I complete my examination of the 1948:2005 RAW data set.

Reply: Anthony’s spam bucket tends to grab stuff with a high link-to-text ratio. I think your post was above whatever that limit is. It looks like one of the moderators did find it and let it get through.

As for what you are seeing, I suspect it is homogenization, TOBS, and FILNET differences.

I’m sure that you know that the daily temperature data from the Mohonk Lake surface station can be found for the interval 05/1948 through 12/2005 at

http://cdiac.ornl.gov/cgi-bin/broker?_PROGRAM=prog.climsite.sas&_SERVICE=default&id=305426

I selected a random sample of ten completed B91 Forms for Mohonk Lake from the site that you referenced

http://www7.ncdc.noaa.gov/IPS/coop/coop.html?_page=2&state=NY&foreign=false&selectedCoopId=305426&_target3=Next+%3E

and verified that the values for TMAX(F) and TMIN(F), with one exception, were identical. The exception which occurred on 07/12/1956 is an obvious transcription error; the value (67°F) listed in the “At OBSN” column was recorded for TMIN(F) instead of the correct value (64°F). In addition to TMAX(F) and TMIN(F), the USHCN site also lists TAVE(F) which equals [TMAX(F) + TMIN(F)]/2 rounded to the next higher integer when either TMAX(F) or TMIN(F) is an odd integer. Therefore, I believe the data archived at the site that I referenced is the “RAW” daily temperature data for station 305426.

So far, so good …

I then calculated the monthly averages for TAVE(F), TMAX(F) and TMIN(F), for the interval 05/1948 through 12/1959 and compared the results with the corresponding monthly values that I downloaded from

http://cdiac.ornl.gov/cgi-bin/broker?_PROGRAM=prog.climsite_monthly.sas&_SERVICE=default&id=305426

How did they compare? Not so good!!

The average differences for the 139 month interval are (s.d. shown in parentheses):

RAW TAVE(F) – HCN TAVE(F) = 1.39°F (0.43°F)
RAW TMAX(F) – HCN TMAX(F) = 1.43°F (0.62°F)
RAW TMIN(F) – HCN TMIN(F) = 0.83°F (1.17°F)

What adjustments do you think were made by NCDC that caused these differences?

I may add to this post after I complete my examination of the 1948:2005 RAW data set.