A new study from USGS by Keven Gallo and George Xian verifies what we’ve already learned and published on via the Surface Stations project; that concrete and asphalt (aka impervious surfaces) have increased near weather stations that are used to monitor climate. In this case, it is the much studied USHCN, that climate network I presented a poster on at AGU 2015. Details here.
What is most important about this paper is that it quantifies the percentage of stations that have had increased amounts of impervious surface area getting closer to the stations. As I have long since maintained, such things act as heat sinks, which increase the night-time temperature when they released the stored energy from the sun that was absorbed during the day as infrared, warming the air near the thermometer, and thus biasing the minimum temperature upwards.
In this study, they have observed over 32% of the USHCN stations exhibited an increase in impervious surface area of ⩾20% between 2001 and 2011. When the 1000 m radius associated with each station was examined, over 52% (over 600) of the stations exhibited an increase in ISA of ⩾20% within at least 1% of the grid cells within that radius.
What this suggests, is that like Las Vegas, which has had huge infrastructure boosts in the last 50 years, that the minimum temperature is creeping upwards, and that biases the mean temperature used to look for the “global warming signal”. NOAA would do well to remove stations that have been encroached upon like this, but they stubbornly hold onto this flawed data, insisting they can “adjust” it to be accurate. I say bollocks to that. Since the USA is so highly over-sampled with thousands of weather stations, it is far better to discard noisy and imperfect data, and use only those stations that have not been biased by infrastructure increases, but retain only the best stations with pristine data.
This is what you get when we did exactly that, and found a statistically significant lower 30 year trend.
Here is the new paper:
Changes in satellite-derived impervious surface area at US historical climatology network stations
Kevin Gallo, George Xian
The difference between 30 m gridded impervious surface area (ISA) between 2001 and 2011 was evaluated within 100 and 1000 m radii of the locations of climate stations that comprise the US Historical Climatology Network. The amount of area associated with observed increases in ISA above specific thresholds was documented for the climate stations. Over 32% of the USHCN stations exhibited an increase in ISA of ⩾20% between 2001 and 2011 for at least 1% of the grid cells within a 100 m radius of the station. However, as the required area associated with ISA change was increased from ⩾1% to ⩾10%, the number of stations that were observed with a ⩾20% increase in ISA between 2001 and 2011 decreased to 113 (9% of stations). When the 1000 m radius associated with each station was examined, over 52% (over 600) of the stations exhibited an increase in ISA of ⩾20% within at least 1% of the grid cells within that radius. However, as the required area associated with ISA change was increased to ⩾10% the number of stations that were observed with a ⩾20% increase in ISA between 2001 and 2011 decreased to 35 (less than 3% of the stations). The gridded ISA data provides an opportunity to characterize the environment around climate stations with a consistently measured indicator of a surface feature. Periodic evaluations of changes in the ISA near the USHCN and other networks of stations are recommended to assure the local environment around the stations has not significantly changed such that observations at the stations may be impacted.
Kevin Gallo is the Corresponding author at: USGS, Earth Observations and Science (EROS) Center, 47914 252nd Street, Sioux Falls, SD 57198-0001, USA.
Note from Anthony: Full disclosure, I was an invited reviewer for this paper, and I submitted reviews that caused improvements (according to the editor) to the paper.