Mount Washington Temperature Trend

In looking at weather stations around the world, one stands out as a fairly unique and potentially representative indicator of temperatures in the lower troposphere for at least higher latitudes of North America and perhaps also fairly relevant for higher latitudes in all of the North Hemisphere.  This unique weather station is located at the top of Mount Washington in New Hampshire in the Northeastern US at an elevation of 1,917 meters (6,288 feet) above “mean sea level” (which is not so easy to define, but that’s another story).

Below is a US Geological Survey topographic map of the area.  Mount Washington is the highest of several high peaks oriented roughly south to north in the Presidential Range.  This orientation may help to orographically increase wind speeds with high westerly or easterly winds.

Topographic map of the Mount Washington area

Topographic map of the Mount Washington area.

Below is a view of the peak without snow.

Below is a winter view with heavy snow cover on the mountain.

Below is a closer look at the round observatory at the top of the mountain.

Below is a photo of the observatory covered in snow.

And a closer look at the observatory coated in rime icing in 2004.

This location is very remote, far away from any urban areas.  However, the temperature measurements could be subject to microscale effects from the nearby observatory.  Frequent strong winds at the site should help to minimize any such microscale effects.

I made a quick check on the internet for annual temperature data for the site and found some available for 1948 through 2015 from Weather Warehouse.  At first I accepted the data, since it was from what I thought was a reputable source.  However, I was a bit suspicious of it because the annual variability seemed larger than expected and the temperature colder than expected.  So I decided to compare the data with what is available from the Weather Underground back to 1973.  The two data sets did not match and the differences were substantial.

I then looked to see if measurements were available from the National Center for Environmental Information (NCEI) and found they had daily data since 1948, but the entire data set was not available for free.  They did however provide a free PDF monthly copy of the daily minimum and maximum temperature observations which included a monthly average.

I went to the trouble to download every monthly PDF copy since 1948, which took awhile, but was free.  I then hand entered (triple checking) all of the monthly maximum and minimum averages into a spreadsheet and then calculated annual averages.  January 1948 was incomplete, but all other months were complete through November 2015.  For December 2015, I used the hourly measurements available from the Weather Underground, including the reported 6-hour maximums and minimums  to compile preliminary daily and monthly data for December 2015 and preliminary data for 2015.  Figure 1 shows the resulting annual average temperatures and associated linear regression trend (click to enlarge).

Mt Washington NH annual temperature trend 1948-2015

Figure 1. Mount Washington NH annual temperature trend 1948 through 2015.

The annual average temperatures show a linear upward trend of +0.0115 degrees Celsius (C) per year, which corresponds to +0.78C over the 68-year period and to +1.15C if it continues for 100 years.

Taking a closer look for patterns in the data, I see three distinct periods as shown in Figure 2.  The first period from 1948 through 1968 shows a statistically robust downward trend of -0.0695C per year, or -1.46C over the 21-year period with a fairly high coefficient of determination (R-square=0.5466).  During this period global carbon dioxide (CO2) concentrations were beginning to accelerate higher from human influence increasing from about 310 parts per billion (ppb) to about 322 ppb based on ice core measurements.  This is an increase of about 4%.  In contrast, Mount Washington temperatures decreased.  The mean annual average for this period was -2.8C.

Mt Washington NH temperature trend patterns 1948-2015

Figure 2. Mount Washington NH temperature trend patterns 1948 through 2015.

For the period 1969 through 1997, the Mount Washington annual average temperatures showed a lot of annual variation, but not much trend.  A linear regression fit for this period indicates a statistically low confidence slight upward trend of +0.0077C per year, or +0.22C over the 29-year period.  The mean annual average temperature for this period was also -2.8C.  During this period, annual average CO2 levels measured at the top of the Mauna Loa volcano in Hawaii rose substantially from 325 ppb to 363 ppb, or an increase of about 12%.

In 1998 there was a large high spike in the annual average temperature that may be related to global effects from the El Niño event that peaked in late 1997 and early 1998 in the tropical central and eastern Pacific Ocean.  For the period 1998 through 2015, the Mount Washington annual average temperatures again show large variations and only a statistically weak trend, in this case slightly downward at -0.0122C per year or -0.22C over the 18-year period.  However, the mean annual average temperature for this most recent period was -2.0C, which is 0.8C higher than the previous two periods.  This change appears to be a step jump that is not consistent with the steady rise in CO2.  The Mauna Loa CO2 annual averages increased from 367 ppb in 1998 to 401 ppb in 2015, or an increase of 9%.

I also looked at the trend in average maximum and average minimum temperatures at Mount Washington as shown in Figure 3.  Trends for both are very similar to the annual average trend.  All of these trends are statistically low confidence, but do show an overall small upward rise across the 1948-2015 period.  Interestingly, the temperatures for the last two years are lower than for the first two years.

Mt Washington NH temperature trends 1948-2015 for maximum, average, and minimum.

Figure 3. Mount Washington NH temperature trends 1948 through 2015 for annual average maximum, annual average, and annual average minimum.

The overall temperature pattern at Mount Washington is very similar during the satellite era to the satellite derived estimates of global lower troposphere temperature patterns and to the Climate Forecast System Reanalysis (CFSR) global temperature patterns described in previous posts.  All of these show evidence of a relatively flat period from 1979 to 1997, then a sudden upward jump apparently associated with the 1997-1998 El Niño, followed by another relatively flat period to the present.  These patterns look nothing like the steadily increasing CO2 concentrations and this discrepancy casts a large measure of doubt over predictions of catastrophic global warming caused by man-made “greenhouse” gases dominated by CO2.  If CO2 was the main driver for global temperatures, the patterns should show a consistent match.  The implication is that CO2 is not the main driver of global temperatures.


2 responses to “Mount Washington Temperature Trend

  1. Congrats on hard work to get the data. I am concerned that your conclusions depend on your choice of break points. If you choose 1978 and 2003 you get a different conclusion. I would think that a running average of a 5 to 10 year window would give you a clearer picture of the evolution of temperatures in Mount Washington, except for the end periods.

    It is reassuring that temperatures at 3 km high in an isolated station match satellite lower troposphere temperature records. Even if it is a single location to global comparison it would be nice if you could show both in the same graph to see how much do they match for the common period. That could be a valuable take home message from your article.

    I agree that the correlation between CO2 increase and temperature increases is built on ad hoc explanations like increased aerosols for the 50’s-70’s big hiatus, or ocean warming for the present hiatus. The alternative, that natural variability plays an important role, is not acceptable to IPCC as it would lead to a rejection of proposed CO2 curbing measures and international agreements as ineffectual.

    Regarding the El Niño effect, it is clear that it confuses temperature records and has a disproportionate effect on satellite and Mount Washington warming, because it is a direct tropical sea surface to atmosphere transfer of energy, so atmosphere warms more than land surface. In my opinion those that believe that El Niño is the main driver of global warming, because the temperature record appears to be flat except for El Niño warming steps, are mistaken. El Niño is a planetary cooling mechanism. It takes energy from the ocean and sends it to space through the atmosphere. It does so when the latitudinal temperature gradient gets too high for the normal oceanic poleward energy transport. It is not the cause of global warming, but a manifestation. A manifestation that distorts the temperature record. It should be ignored, and we should never choose breakpoints before or after a big El Niño or La Niña.


  2. Javier, thanks for your support. I agree that diagnosing patterns in the data can be somewhat arbitrary and subject to political agendas. I have looked at lots of various kinds of data over many years now, so I am used to looking for patterns in data. The patterns I reported are the dominant patterns that I see, but everyone has a different view.

    I had thought about graphing the Mt Washington and UAH global temperature anomalies together but the post was already getting long. I will take a look at it for a future post. The main point is that I see the same pattern in both, with flat trends in the satellite era disjointed by an step up offset at the Great El Niño of 1997-1998. Since the current El Niño appears to be close in strength, it will be interesting to see if there is another step change, but we may have to wait several years to be sure. It will also be interesting to see how much of a La Niña develops later this year. The current El Niño seems to be fading fast judging by the latest SST anomalies.

    I agree that the ENSO is not likely to be a driver of climate, but more of a symptom and player. I do not have a hypothesis as to why the global temperature made an upward step after 1997-1998. It may be pure coincidence that other factors caused this step change and not the process related to the Great El Niño. If the step change was not related to the process of an intense El Niño, then we may not see another step change this time.

    The whole climate process is like the chicken and egg problem. What comes first when there are never ending cycles affected by many natural and a few man-made factors? The process is so complicated and chaotic that I am doubtful humans will ever be able to develop reliable long-range global and regional climate models. To be reliable and useful, these models must be able to consistently make better predictions than can be obtained by using either persistence or past climatology. If these models cannot consistently make accurate short-term climate predictions over months or a few years, why should we expect them to be accurate for longer periods?

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 )

Twitter picture

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

Facebook photo

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

Google+ photo

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

Connecting to %s