The Blended Temperature Set

I have previously discussed the temperature set that I have put together for my usage here.  It has been updated a bit since then and people have been asking questions so I guess this would be a good time to provide an update and compare it to other temperature sets.  It is important that people understand what the temperature set I use is and where it comes from.  I have always been transparent about my usage of the blended temperature set.

There are two reasons I chose this particular path.  The first is that it is scientifically proper to chose one temperature source and then stick to it.  Each of the different sets has different benefits and problems.  Often times it is possible to determine the conclusion of a persons argument based on the temperature set they chose to use.  A person that is using the UAH set is probable to be a skeptic.  An argument using the GISS temperature set is likely to be a warmist.  I consider that simple bias as a good reason to go a different route and stick to it.

The second reason is that each set is a different method of interpreting the available information.  In what I would consider a best engineering option I decided to merge 4 sets into a single set.  By choosing the duration of records from the CRU and GHCN (initially HadCRU, but I switched later) with the high resolution of the UAH and RSS modern satellite sets I decided to make a single temperature set that I would use regardless of how well it fit with what I expected.  I have stuck to that except when I was discussing the behavior of a particular set.  This has been important as the UAH, RSS and GHCN have tweaked themselves over the past few months.

One thing I noticed early on is that the HadCRU was a reasonably good match for the blended set.  The comparison for the two sets is shown below.  I am aware that CRU is a component of the HadCRU, but once past 1880 it is not a significant portion of the results.  Once past 1979 which is when the main concern of global warming began there is even less of a dependence as the satellite data came on-line.

The Inconvenient Skeptic

The full range of the blended temperature set compared directly to the HadCRUv3 temperature set.

The error for the two sets is 0.084 °C.  That is comparable to any two other temperature sets.  A look at the same RMSD between temperature sets gives the following results:

CRU vs.  GHCN:        0.120°C

CRU vs HadCRU:       0.094°C

RSS vs UAH:            0.097°C

HadCRU vs GHCN      0.138°C

In all cases the error is less between the blended set that I use and between other sets.  My method for creating this set may be unorthodox, but I have created a temperature set that incorporates 4 different sources of temperature data and have achieved a result that is reasonable and usable.  It could be argued that I should use the HadCRU data, but then I would be back to using a single temperature set.  The purpose was to incorporate as much data as possible into a single set.  It would appear that because the different sets use different periods for setting the anomaly that in the end I have a more robust data set that is less likely to be influenced by any particular bias that a single set might have.

Here is the comparison of the two sets since 1970-present.

The Inconvenient Skeptic

The two sets show similar behavior for the past 40 years, but the blended set shows more response in individual years as expressed by it's higher variability.

Events like the eruption of Mt. Pinatubo or the ENSO cycle show up more clearly in the blended set than they do in the HadCRU set.  This is because of the increased sensitivity and more comprehensive coverage of the included satellite temperature data.

A comparison of the four components that I use to create the blended shows why different groups tend to prefer one to another.

The Inconvenient Skeptic

When the individual components of the blended set are shown together it is clear why different sets get chosen to make a point. There are significant differences between the station and satellite temperature sets.

The two most extreme differences by set are the UAH and CRU.  The error between them is an astonishing 0.381°C.  Since the UAH data covers the entire Earth from 85S to 85N it is the highest resolution temperature set available with resolution as good as the RSS set.  Ignoring that data is foolish, but using it exclusively when it only goes back to 1979 is also folly.  The blended temperature set is my method.

I also know that each data set is used accurately from each source.  That is why I keep finding tweaks to the different sets.  They show up even if nothing is announced.  It also highlights differences when one set responds differently than the other sources.  This happens more often than I expected which is another reason why I trust this set.  Finally I would like to show it by itself.  It shows that the Earth has warmed about 0.7 °C in the past 100 years.

The Inconvenient Skeptic

The full range of the blended set by itself. The multi-set data shows that warming has happened.

Posted in Anomaly and Skeptic by inconvenientskeptic on March 16th, 2011 at 1:28 am.


This post has 22 comments

  1. Joris Vanderborght Mar 16th 2011

    The urban heat island effect is something that should be taken in account. It would be interesting to compare measurements in areas where there is no chance that the UHIF has had any effect with the complete body of measurements. Even smaller concentrations of buildings can have an effect (for instance an airfield modernising over the time). So i would not consider measurements at airfields as measurements which are free of the UHIF.
    Could this effect explain a part of the warming?

  2. Sunshine Hours in many locations have also risen in the 20th Century. Global Brightening occurred in 1920-1940 period, dimming occurred in 1960-1990 period and brightening occurred again in the 1990+ period.

    An increase in sunshine that reaches the ground can explain some (if not all) warming.

  3. SoundOff Mar 16th 2011

    All five major temperature data sets report the same overall trends even if the monthly and regional anomalies differ slightly. Satellite data sets are a bit spikier due to their greater sensitivity to ENSO episodes, which move large amounts of heat upwards by convection, and to volcanoes. Otherwise there is little difference.

    See this link for an expert statistician’s comparison of the datasets.

    And see figure 7:

    See the next comment for a backgrounder I wrote about satellite temperatures.

  4. SoundOff Mar 16th 2011

    Satellite Temperatures

    I put a little less faith in satellite temperatures than surface measurements though they are still quite good. Here are my reasons why.

    A temperature satellite doesn’t really measure temperature. It measures radiation in certain wavelengths, specifically very LW microwave radiation coming from oxygen molecules throughout the atmosphere. This radiation is proportional to the temperature of the molecules, so temperature can be inferred from the radiation. But it’s not as clean and simple as a surface thermometer that operates on thermal expansion.

    Satellites give a cross-section or average of the radiation received from the surface up to the top of the atmosphere. These readings aren’t really surface temperatures but they are suitable for examining climate trends if one assumes the atmosphere tracks the surface. But the atmosphere doesn’t exactly track the surface. The atmosphere warms/cools quickly compared to the surface air over oceans and the land areas near them (over 70% of the planet). Satellite records show stronger up and down spikes in temperature in response to El Niños and volcanoes. But the trends are similar.

    Satellites don’t go around the Earth in an exact circle. They rise and fall as they pass over large gravitation masses like continents, mountains, glaciers, etc. They also drift laterally from their theoretical orbits (due I think to the spin of the Earth, atmospheric drag, lunar tugs, etc.), which means they aren’t always reading what we think they are reading. Finally, orbits slowly decay over time. All these movements mean their radiation readings can’t be used at their face value; they need much error correction. Just to further complicate things, I think when they pass over cool clouds and warm cities they can get false readings too. Another thing is a polar orbit is unstable so satellites can’t fly directly over the poles and that leaves a geographic hole in their measurements.

    A lot of mathematics is needed to make the necessary corrections. Two organizations regularly process satellite readings to arrive at temperatures – Remote Sensing Systems (RSS) and the University of Alabama in Huntsville (UAH). They use different techniques so they arrive at slightly different answers, not unlike their surface reporting equivalents, NASA-GISS, CRU, NOAA, JMA, etc. Of course various scientists put their own twist on the readings in many one-off studies. Satellite temperature records started around 1979, so their record is not very long for climate analysis, which needs 30 years of data for reasonable trends. Originally, satellite records showed cooling while climate models and surface records showed warming. Satellites were the favourite reference of skeptics who claimed the models were wrong and that the surface record was corrupted. Then in 2004 balloon observations were able to prove that there were systematic errors in the satellite math (some exotic drift or decay was being missed) and now satellite temperature records agree with other measurements with the further correction.

  5. intrepid_wanders Mar 16th 2011

    SoundOff – I am not trying to diminish or inflate your points, but you are going into infinite possibilities that have no specific answers. A simple question of how time dilation (in a 1 to 1 relation, no problem, but with the theory of CO2 bouncing and collecting, this can be another factor) can affect satellite readings would be worthy of a study. If there is a transformation of the wavelength, there could be a futher transformation based off of a gravity induced “forcing” in a time-space concept (LW -> Gravity Object, SW <- Gravity Object). With the RSS, new unknowns are introduced.

    It is unusually the math that is problem, typically the analysist 😉

  6. Soundoff, do you know much off the SST records were compiled by sailors throwing buckets into the ocean, and then taking old fashioned thermometers and dipping them int he bucket?

    Some were metal which caused the water to change temperature and some sithced to plastic buckets, then some switched to inlet water ….

    Supposedly 70% of the earth has “accurate” measurement taken by the occasional ship traversing the same few sea lanes.

    Your criticism of satellite temperature may be accurate, but SST and land temperatures are 10-1000x less accurate. I’m guessing. But so were the sailors.

  7. SoundOff Mar 17th 2011

    Bruce, I’m not really criticizing satellite temperature records. I’m just noting that they have their own risks, peculiarities and biases. I’m confident that they are more-or-less correct. UAH numbers are a bit more suspect in my eyes since they are the only outlier among all the temperature records (UAH shows a bit less warming), and UAH has a history of prior mathematical errors. Combining the records into a single data set can be done but it must be done carefully (adjustment to a common base period is critical) and the result makes it somewhat difficult to explain effects as different kinds of noise can exaggerate the trends at times and then cancel out at other times. It might just make the long term signal noisier.

    Yes, the northern hemisphere SST component of the surface temperature record during the 1940s is also a bit suspicious. There’s a temperature bump then that doesn’t show up to the same extent in any of the land records or at all in the southern hemisphere SST record. Likely this was caused by the switch from cooler bucket measurements to warmer engine intake measurements. It’s being studied. There’s little doubt about any other part of the surface temperature record for the last 100+ years. After all, mercury thermometers have been in use for almost 300 years (and a couple centuries longer for other types). Satellite thermometers are still in their infancy, or just hitting school age.

  8. inconvenientskeptic Mar 17th 2011


    I have two separate responses to the satellite issue. First is that nothing actually measures temperature. Station data is typically a two metal thermocouple (TC) that emits the voltage as the thermal expansion of the metals change with the temperature. They can be very accurate, but they are much like that satellite in that they need to be calibrated.

    That the satellites measure microwave feedback in the EM spectrum makes them comparable to another method of measuring temperature called a pyrometer. Those are used for higher temperature because of the wavelengths associated with room temperature are not very energetic. Microwave works well for room temperatures and that is why microwaves are used.

    Both types of measurements require careful calibration and math to convert what is measured into a readable temperature. That a TC generally has all of that built into the device is irrelevant. It is still a delicate device that does not directly measure temperature. Of the two methods the microwave one is the one that is actually closest to measuring temperature as the results are direct EM measurement.

    The 2nd response deals with analyzing the response of the two temperature methods to climate events. You refer to the greater sensitivity to them, but that would indicate that the satellite measurement system should be more responsive to any type of warming or cooling. Here are some actual differences in response to the types of easily detectable events like a volcanic eruption or ENSO events. For instance after the Mt. Pinatubo eruption the UAH and RSS dropped globally by 0.285 C. The CRU and GHCN dropped by an average of 0.174C. More than a tenth of a degree difference in measurement sensitivity. The satellites detected the cooling effects more clearly than the station sets did.

    The warming effect of the 1998 El Nino stood out more in the satellite data. The station sets show an average global increase of 0.224 C. Over the same time period the satellites showed a 0.46C response to the warming. The satellites detected 105% more response of the Earth’s climate to the 1998 event. If detecting a change is the purpose of monitoring the climate, the satellite measurement is clearly and statistically superior at detecting both warming and cooling events.

    Since the satellites are demonstrably superior at detecting changes in the Earth’s surface, why do you have less faith in them? Your argument points to their complexity, but the results show that they are distinctly better at detecting temperature changes. To quote Darth Vader, “I find your lack of faith disturbing.”

  9. Sound, Mercury thermometers are only accurate to 1F, which happens to be about what some people have claimed is CO2 caused warming.

    “Yes, the northern hemisphere SST component of the surface temperature record during the 1940s is also a bit suspicious.”

    No, all SST measurements are suspicious. All land measurements are slightly less suspicious if you read anything about Anthony Watt’s surface station project.

  10. inconvenientskeptic Mar 17th 2011


    Since I incorporate many sources I also get the different groups attempts to deal with the urban heat effect. One distinct advantage in merging the different sources is that if there is a positive bias in two and a negative bias in the other two. The end result will be neutral.

    The main problem that the blended set would have is if all 4 groups had the same bias in one respect, but there is no solution around that unless I caught the specific bias.

    I am pretty happy with the results. I just finished getting a monthly time series done for the entire period. It is interesting.

  11. SoundOff Mar 17th 2011


    1. Every temperature measuring instrument needs calibration, thermocouples and even mercury thermometers. However, we can’t put some other kind thermometer that we trust next to a satellite to validate that the satellite is measuring correctly, which we can do with any other type of instrument. Radiosondes are our closest option, and when we did that last, we found that the satellites were wrong. Fortunately, the satellites readings now generally agree with surface-based records, so I guess they should all be equally suspicious to Bruce now. (BTW Bruce, the weather stations in my country measure in tenths of a degree Celsius, but I’m not sure about the USA where they still use those ancient pint-sized Fahrenheit degrees. News reports here generally round to the nearest degree.)

    2. Satellites are not demonstrably superior at detecting changes in the Earth’s surface because they are detecting changes in the atmosphere instead, at least mostly. That’s why they are more sensitive to ENSO, volcanoes, etc. That’s fine if that’s what you want to measure – they do that well. But most of us live on the surface. So some hopefully smart people need to infer surface temperatures from the satellite readings. That makes it more of an art than a science.

    3. I really don’t see the value of combining different temperature records. Group A sums all the T measurements from each grid box and divides by the number of boxes to give a global average. Group B sums the same T measurements but from each NH or SH grid box separately and divides by the number of hemisphere boxes, then adds the two resulting numbers and divides by 2 to give a global average. They you come along and average their differently determined averages yet again. It’s just average on top of average on top of average. Nothing is gained by doing it. And when you mix a surface T with a satellite T it can even get weird depending on how carefully it was done.

  12. inconvenientskeptic Mar 17th 2011


    You make good points. I feel that incorporating more data is better. I also trust that the coverage provided by satellites is now superior to what stations can provide by a significant margin. That has been seen by the greatly improved weather forecasts.

    My method might be unorthodox, but does it appear to be reliable? The comparison to HadCRU is pretty good. Certainly my results appear to be a better match than the satellite to stations are to each other.

    If everyone agreed to a single set, I would use it. Since people pick the one they want now, I am meeting each side halfway.

    Even if you disagree with the method, do you see any problem with the results?

  13. SoundOff Mar 18th 2011

    TIS, I have no problem with you using any single set of your choice since they all say essentially the same thing. No particular set is more right. Your blended set is just an attempt to track somewhere between the others. As long as you made the critical adjustment to a common base period before combining the different data sets and you took into account the different periods that data is available for, then your result is valid, though not necessarily an improvement in my opinion.

    If you don’t make the critical base period adjustment, then you are doing the equivalent of averaging dollars, euros, rubles, pesos, dinars, yen, yuan and beaver pelts to come up with an average income figure across several countries. The base period adjustment is quite difficult to do in a reliable way when you don’t have a blended exchange rate to apply to data in the periods where there’s no overlap (i.e. any time before satellites).

    Improved weather forecasting is more a result of other kinds of satellites that track clouds and storms than from global temperature measuring satellites, which are mainly used for ocean temperatures where no fixed stations exists. I think they even use a different kind of temperature measuring satellite for weather purposes (called MODIS, which has only been available since 2000 – I believe these are geo-stationary satellites that can see through clouds) but weather forecasting is all outside my area of study, so I could be wrong.

    I stumbled upon this interesting and relevant snippet in Wikipedia:

    However, there are several difficulties with satellite-based absolute SST measurements. First, in infrared remote sensing methodology the radiation emanates from the top “skin” of the ocean, approximately the top 0.01 mm or less, which may not represent the bulk temperature of the upper meter of ocean due primarily to effects of solar surface heating during the daytime, reflected radiation, as well as sensible heat loss and surface evaporation. All these factors make it somewhat difficult to compare satellite data to measurements from buoys or shipboard methods, complicating ground truth efforts. Secondly, the satellite cannot look through clouds, creating a cool bias in satellite-derived SSTs within cloudy areas. However, passive microwave techniques can accurately measure SST and “see” through clouds. Within atmospheric sounder channels on weather satellites, which peak just above the ocean’s surface, knowledge of the sea surface temperature is important to their calibration.

  14. Richard Mar 18th 2011

    Trying to compare two methods of temperature sensing to derive an whole world average (one a point method and the other an area method) is always going to be challenging.

    Although the ground based thermometers may be more precise about the temperature at a given point they can only give an statistical approximation to the temperature over an area. Likewise the satellites are probably more precise about whole area temperatures but more vague about particular points within it.

    In a ideal world we would have reliable temperature records by either method that went back 500 years. Then we might have some real data on what the actual trends or cycles are.

    Otherwise way too much quesswork becomes involved (often hidden as complex statistics).

  15. Richard Mar 19th 2011

    I think that possibly one of the most valuable views of all this would be to plot vertical atmospheric contour plot cross sections on a Longtitude basis.

    These plots could show the vertical air temperature cross section with the instrument temps at the surface (if required) and the rest either from ballons or satellites from 180S to 180N.

    These can also be extended to show ground and sea floor elevations as well as water temperature at depth if required.

    Also comes the possibilty to add in clouds at the correct position vertically.

    As these sections are at right angles to the main energy flows in the system they should also well demonstrate the previously known patterns in both air and water.

    After all it is really only the transfer of energy from the equator to the poles (and any changes in time) that we want to understand.

    Course or fine details can be obtained by increasing the number of sections from 36 to 360 to 3600 or more.

    This gives the other dimension to the vertically viewed plots that we are all familiar with and may provide other valuable insights.

    If extended to pressure and mass density (both of which directly effect temperature change) this may give a better overall view of what is happening.

    What do you think?

  16. SoundOff Mar 19th 2011

    Some of what Richard wants in his prior comment is covered at the link below. It’s hard to show this graphically unless limited to one site and time because the ambient and adiabatic lapse rates vary widely depending on which latitude of the Earth is examined as well as the temperature and humidity of the local air mass, and, of course altitude (air pressure).


    “from 180S to 180N” ???

  17. Richard Mar 20th 2011

    Sorry, I obviously was not clear enough. What I was thinking about was a contour plot of temperature in the vertical direction for the whole of a single Longtitude from 90S to 90N. Like the existing temperature contour plots that are in a horizontal direction but in a vertical direction instead.

    Time wise this would be monthly averages to reduce the ‘noise’ of weather.

    This format would allow for both land and sea elevations to be displayed as well as actual atmosphere layers. Vertical and horizontal velocities (in a North-South plane) of both air and water at different heights are also easy to display.

    As I mentioned before, this cross section is at right angles to the main energy flow in the system from the equator to the poles as well as correctly encapsulating the warm air sitting below the colder air above as the RClimate page shows (but in a different layout).

    I can find plots for solar input that run from 90S to 90N (and how that changes over the year) and there are multiple different altitude temperature plots in the horizontal direction of a similar range but I can find no contour plots in the vertical direction.

    This is slightly strange because I believe that it is this very direction that any climate modification or change would be easiest to see.

    The other question that becomes relevant would be ‘how many sections’ but that is the same problem as with any other gridding system.

  18. SoundOff Mar 20th 2011

    Richard wants a one month average temperature contour plot by altitude and latitude for one entire longitude, with clouds, surface elevations, wind velocities/directions (vertical and horizontal) and all atmospheric layers shown. He’s asking for a lot from one 2D plot.

    I think what Richard wants is partially do-able but I’m not aware of it having been done already. Detailed vertical profile data are only available from certain weather stations that have radiosondes to measure air temperatures at many levels. Satellites don’t have that kind of resolution; they give a smeared result.

    What Richard needs to do is pick a longitude that has many stations and line up all those stations north to south. There will be many gaps unless he repeats the process for many longitudes and then averages them all together (there will still be little data over oceans). Then for each station, visit the link similar to the first one below. Do this for each day of one month to get all the detail data (there are huge seasonal shifts so maybe a whole year or even several years should be used). Next extract the temperature reading for each altitude (the some station readings only go to 13 km, others to 35 km, I’m not sure why). Then average all those daily altitude-temperature readings by station latitude. I suppose the latitude gaps could be filled in with values extrapolated from a line or curve between the known values so that the contour map does not have too many empty spots. Lastly put all this data together and plot it in a form similar to the second link below except with altitude being on the x-axis.

    I doubt if the end result will tell us much. If red is warm and blue is cool, the chart would show a strong red hue on the left (low altitude) and would be shaded through to strong blue in the center (mid altitude) with a very slight shift back to red on the right side (high altitude). The red-blue transitions would be very wavy according to the conditions at the specific station(s) selected and the month examined.

    I don’t know how to represent wind velocity/direction or cloudiness by altitude on this plot or where to find this data by station. Elevation is available in the detail data link but it will be scarcely visible if Richard plots it on the altitude scale. I’m not sure what Richard means by displaying â��atmosphere layersâ��. All the available data is for the first two layers called the troposphere (first ~10 km, the weather and clouds layer) and for the stratosphere (next ~40 km, the layer with high flying jets and where ozone is at work). Ongoing measurements are not available for higher layers nor are they relevant to climate.

    Sample of detail data for Oakland, CA:

    Contour map � go to: and click show map.

    Temperature profile of various layers:

    There are some useful things here:
    There’s lot of nice short animations at:

    Richard, please publish your plot here when you have it done. I’m eager to see the result.

  19. SoundOff Mar 20th 2011

    Since Richard is interested in wind directions, perhaps this will be helpful.

    Thinking horizontally, surface wind directions are generally west-to-east or vice versa depending on the specific latitude, though they are angled north-south one way or the other within each band. There are three bands of wind direction in each hemisphere � tropical trade winds blow from the east converging on the equator (ITCZ), mid latitude westerlies blow from the west converging on the Arctic Circle and polar easterlies blow from the east converging on the Arctic Circle as well but from the opposite direction. These directional differences explain why hurricanes head west in the Caribbean and then turn and head east after they hit landfall further north over the USA. The bands shift slightly north and south with the seasons (causing dry and wet seasons in some areas). The boundary between the bands is not a straight line (like latitudes). Rather they are very big wavy lines that shift north or south over the continental land masses seasonally and shift less over oceans. Higher up, above 5 km of altitude, winds generally move west to east regardless of latitude. See

    Thinking vertically, warm tropical air rises to the top of the troposphere (creating a low pressure zone causing much rain). Then it flows towards the poles at higher altitudes. Meanwhile, the cool air at the poles sinks (creating a constant high pressure zone resulting in a desert). This cool air then flows towards the equator along the surface. All flows have a westward shift due the Coriolis Effect. See My simplified description isn’t quite true as the vertical circulation actually stays mostly inside a single band within â��cellsâ�� separated by â��jet streamsâ�� and the flow is reversed in the mid latitudes where the westerlies blow (see the first link above again or for much greater detail). The vertical convergence of the normal tropical flow and reversed mid latitude flow causes a constant high pressure zone between the two so we have a band of deserts around the world in that area. This partial isolation of air flows within cells means the atmosphere is not a global heat transport engine in the fullest sense.

    Of course, both of the above effects are often mixed up and sometimes overwhelmed by other factors like land/ocean variations, mountain/lowland magnification and especially local highs and lows (i.e. weather). Thinking vertically, an ocean surface warms up causing water to evaporate, warm air rises, clouds form and the clouds move over land shading a large area. That area becomes cooler and the air becomes heavier so it sinks (and the sky clears because the clouds dissolve as the cool air sinks and warms up). Meanwhile, warmer air in a nearby sunny area rises due to its relatively lower pressure (usually creating clouds and rain as the rising air cools). Thinking horizontally, air pushes away from the center of the high pressure area and is pulled into the center of the low pressure area. This happens on a spinning planet so the wind actually circulates around (in opposite directions) the center of the high or low but it is generally moving from the high to the low area. The wind circulation is uneven due to air friction with the surface, so the circulation moves in spirals of uneven strength that we see as gusts, etc.

    I believe the main north-south transfer of heat comes from ocean currents which retain their heat over long distances and periods. Like the atmosphere, the oceans heat up and rise at the equator, flow across the surface, sink near the poles and return via currents at depths to the equator to start the cycle again. A cycle takes many decades. This is the great ocean conveyor belt. The prevailing surface winds described earlier push the conveyor belt along specific paths around the land masses. See or

  20. SoundOff Mar 20th 2011

    Hmmm. This site did not like any apostrophes (‘), double quotes (“ & ”), dashes (–) and ellipsis (…) in my earlier comments today. Not sure why. Let me know if I should post the comments again.

  21. inconvenientskeptic Mar 20th 2011

    Not sure why that happened. I have not seen that issue before. Your other post is ok as well. ????

    When it was clear what the text error was I tried to clean it up. If I wasn’t sure I left it… garbled.

  22. Richard Mar 21st 2011


    The plot I was thinking of is

    South – Left 180S
    North – Right 180N
    Up – Vertical elevation
    Down – Vertical depth
    0 = Sea level

    Contour – Temperature in degrees.

    That would make the warm air temperatures a blob in the middle with cold air above and cold water/ground below.

    I accept that putting point sampling ground stations actually on the particular Longtitude is a matter of some form of interpolation from their actual position but that is done elsewhere as well in other griddding systems.

    The point I am trying to make is that that this plot is at right angles to all the major energy flows, both N-S and Up-Down, which are in the directions of the energy travel in the system.

Web Design & Dev by

Mazal Simantov Digital Creativity