Urbanization effects on GHCN temperature trends, Part II: Evidence for abrupt warming homogenization

news7g02/08/2023

3 5 minutes read

From Dr. Roy Spencer’s Global Warming Blog

by Roy W. Spencer, Ph. D.

IN my part I have shown Landsat satellite-based urbanization measurements around the Global Historic Climate Network (GHCN) soil temperature monitoring stations. Almost all GHCN stations have seen an increase in the coverage of human settlement “building” (BU) structure.

As an example of this growth, here is the 40-year change of BU values (ranging from 0 to 100%) at 1 km spatial resolution over the Southeastern United States.

Figure 1. Change over 40 years of urbanization in the Southeastern United States from 1975 to 2014.

How has this urbanization change been demonstrated at the GHCN stations distributed in the world? Figure 2 shows that urbanization increased on average over 19,885 GHCN stations from latitudes 20N to 82.5N, at different spatial average resolutions of the data.

Figure 2. Forty-year mean change (1975 to 2014) in Landsat-based urbanization value (BU) across 19,885 GHCN stations from 20N to 82.5N at five different averages of data 1 km BU.

None of the 19,885 GHCN stations experienced negative growth, which is not surprising since that would require the removal of human settlement structures over time. In all subsequent analyses, I will use the BU’s average 21×21 km centered on GHCN station locations.

So, what effect does urbanization measured in this way have on GHCN temperatures? And, in particular, about the temperature trends used to track global warming?

Although we all know that urban areas are warmer than rural areas, especially at night and during summer, does increased urbanization lead to artificial warming at GHCN stations that have undergone development (most of them)?

And, even if it did, Does the homogenization process NOAA uses to correct for spurious temperature effects cancel out (even partially) the urban heat island (UHI) effect on reported temperature trends?

John Christy and I tested these questions by comparing GHCN temperature datasets (both unadjusted and adjusted). [homogenized] version) for Landsat-based measurements of human settlement structures, which I will just refer to as “urbanization”.

This is what I am looking for so far.

The strongest warming of UHI along with urbanization growth occurs in stations near rural areas

EQUAL Okay (1973) and others have demonstrated, the urban heat island effect is strongly nonlinear, with (for example) a 2% increase in urbanization in rural areas producing more warming than a 2 % in urban areas. This means that climate monitoring datasets using mainly rural stations are not immune to artificial warming from escalating urbanization, unless there is absolutely zero growth..

For example, Figure 3 shows the sensitivity of the GHCN temperature (absolute) to increasing urbanization in different types of urbanization, based on more than 1 million station pairs less than 150 km apart. .

Figure 3. Average barrel temperature change calculated with change in urbanization (BU), in average bins of 2 BU stations is 0-2%, 2-5%, 5 -10%, 10-20%, 20-30%, 30-40%, 40-50%, 50-60%, 60-70% and 70-100%, for four seasons and all GHCN stations in the range latitude 30N-70N. Solid lines are for adjusted (homogenized) GHCN data and dashed lines are for unadjusted data.

By far, the greatest sensitivity to the change in urbanization in Figure 3 is at 0-2% (near rural). We also see in Figure 3 that the homogenization procedure used by NOAA only reduces this effect by an average of only 9% in all seasons and even less (2.1%) during summer.

If we integrate the sensitivity in Figure 3 from zero to 100% urbanization, we get the total UHI effect on temperature (Figure 4).

Figure 4. Seasonal average UHI effect across all GHCN stations between 30N and 70N by integrating dT/dBU values in Figure 3 from 0% to 100%, for temperature data (copper) best) adjusted (homogeneous) and unadjusted data (dashed). The black curve is an exponential relationship with the square root increase in temperature of urbanization.

The temperature data used here is the average of the daily maximum and minimum temperatures ([Tmax+Tmin]/2) and since most urban heat island effects are expressed in Tmin, the temperature scale in Figure 4 would almost double for the Tmin UHI effect.

The black curve in Figure 4 is a square root relationship, which seems to fit the data reasonably well for most GHCN stations (usually with urbanization rates below 30%). But this is not nearly as nonlinear as the 4th root relationship that Oke (1973) calculates for some eastern Canada stations, using population data as a measure of urbanization.

But what I have shown so far is based on space information (difference between stations near each other). It doesn’t tell us whether pseudo-warming exists or how much in the GHCN trend. To examine this question, I next looked at how NOAA’s homogenization process changed station trends as the station environment became more urbanized.

The homogeneity of NOAA produces variation in most of the station’s temperature trends. If I calculate the average homogenization-induced change in trends in different types of station growth during urbanization, we see a negative trend correction related to growth. active urbanization, right?

But the opposite happens.

First, consider what happens at stations without urbanization. In Figure 5, we see that 881 stations with no urbanization trend for the period 1975-2014 tended to be 0.011 C/decade warmer on average in the adjusted (homogeneous) data than in the unurbanized data. adjust. This in itself is quite possible due to the adjustments in observation time (“Tobs”) made to the data, adjustments for station movement, device type, etc.

Figure 5. Adjusting the temperature trends of the GHCN station from the homogenization process inexplicably increases the station’s temperature trends as urbanization progresses, instead of decreasing them as would be expected if the homogenization process The homogenization of NOAA eliminates spurious warming due to the urban heat island effect.

So let’s assume that the value at zero growth in Figure 5 represents what we should expect for non-urbanization-related adjustments to the trend toward GHCN. As we move to the right from zero urbanization growth rates in Figure 5, stations with increasing urbanization should correct the temperature trend downwards, but instead we see , for all types of urbanization growth, adjust UP instead!

Thus, it appears that NOAA’s homogenization process is artificially warming (average) station temperature trends when it should be cooling them. I don’t know how to conclude any different.

Why are NOAA adjustments going in the wrong direction? I do not know.

To say the least, I found these results… curious.

OK, so how big is this artificial warming effect on soil temperature trends in the GHCN dataset?

Before you jump to the conclusion that the GHCN temperature trend has too much of an artificial warming to be relied on to track global warming, what I have shown doesn’t tell us the temperature trend. how much is the mean on the ground deviated. I will address that in Part III.

My preliminary calculations to date (using the UHI curves in Figure 4 applied to the 21×21 km urbanization growth curve in Figure 2) show an average UHI warming across all regions. station is about 10-20% trend of GHCN. Small, but insignificant. But that may change as I dig deeper into the matter.

4.8
4
votes

Rate Articles