Nights warm faster than days.

When we are told that temperatures have risen by 1.2C since  pre-industrial times, most of us assume this simple means that daytime temperatures are increasing. So summers are getting hotter and droughts are getting more extreme, but this is not true. The global temperature data that climate science relies upon  is all based on daily measurements across thousands of weather stations, some going back to the  1700’s.  All these stations record the average temperature for each day based on the minimum (night-time) and maximum (day-time) temperatures. Originally these were measured by physical Min-Max thermometers, whereas today they are all automated digital recordings. The daily average temperature though is always simply

Tav = (Tmin + Tmax) / 2

The monthly average of <Tav> is calculated for each 12 months of the year and  a  30 year climatology derived for each station by averaging <TAV> over 30 years.  These are called “normals” and the 30 year period a “baseline”.  CRU uses a  1961-1990 baseline and NASA uses a 1951 – 1980 baseline.  Finally all the monthly temperatures at each station are subtracted from the monthly “normals” to give monthly temperature “anomalies”.

All results are based on temperature “anomalies” which are averaged over the earth’s surface to give annual temperature series relative to the chosen baseline. These series then show a net warming of roughly 1.2C  since pre-industrial.

Comparison of major temperature datasets (combined land and SST)

However they don’t tell you how that warming is happening on land. To investigate that process we must return to the daily values. GHCN-Daily is a huge archive (33600 stations) of measurements dating back to 1750. Each station recorded  the maximum temperature and minimum temperature each day of operation. Essentially this also measures the diurnal temperature range because Tmin always occurs at night and Tmax occurs after midday.

Tav ( average Daily temperature) = (Tmax+Tmin)/2

Trange (Daily Diurnal Range) = Tmax – Tmin

These are the results I find for the average temperature and the temperature range analysing GHCN Daily

Temperature and Trange anomalies for full time range- Annual averages.

The temperature range has fallen since 1880 whereas the average has increased. This means that minimum temperatures (night)  are increasing faster than maximum temperatures (day). Berkeley Earth gives a similar result

Tav and Trange anomalies (Berkeley Earth)

We can estimate the changes in Tmax and Tmin from these results. The average land temperature has risen by ~ 1.7C since 1880 whereas the range has fallen by ~ 0.7C so it appears that on land maximum temperatures have risen by ~1C since preindustrial times whereas minimum temperatures have risen by ~ 1.7C

This corresponds to our everyday experience. The UK sees fewer frosty days than it did in the past and in my experience there is less overnight snow than there was say 40 years ago. Night-time temperatures in my experience are warmer but the same cannot really be said about maximum temperatures. There have been no heatwaves to match that say of 1976. However the UK is affected by a maritime climate but we can see an even  stronger effect in an arid continental climate like Australia.

Average maximum land temperatures in Australia compared to average minimum temperatures

Annual maximum temperatures across Australia have hardly changed over the last 100 years whereas it is annual minimum temperatures that have risen by over 2 degrees. This explains the observed rise in average temperatures.

So why should the minimum temperatures (night, winter) rise faster then maximum temperatures? The greenhouse effect provides an explanation. During daytime in summer with little cloud, solar directly radiation heats the surface. This initiates strong convection currents which transport heat up through the atmosphere more efficiently than radiative transfer. Increasing CO2 increases the radiative emission height.  However at night solar radiation is zero and the surface mostly cools by radiation only. That is also what causes dew and frost in winter.

Figure from Richard Lindzen. Pure radiative equilibrium would be the temperature gradient without convection. The surface temperature would be >20C warmer than today ! Thermodynamics drives the lapse rate towards the moist adiabatic lapse rate

As a result increasing CO2 increases the night-time temperatures more than the midday temperatures. This reduces the diurnal temperature difference by warming minimum temperatures faster than maximum temperatures. So the first noticeable effect of global warming is less severe winters rather than more heatwaves in summer. So it is  winter sports tourism which suffers more than the summer holiday industry.  I would argue that this has already happened. There is no doubt that we have less snow and ice in the UK than what I remember 60 years ago, and the ski season in Europe has certainly got shorter with lower lying resorts suffering the most.

Posted in AGW, Climate Change, climate science | 21 Comments

Global Temperature update

The latest HadSST4 data release allows me to calculate global temperatures up to March 2022 based on spherical triangulation. HadSST4 results in higher temperatures compared to HadSST3 relative to the baseline of 1961-1990. I combine this with GHCN-V4C station data. The results for the monthly temperatures are shown below. There is still evidence of a weak La Nina and cool spots in Eastern Europe and central Siberia.

Temperature anomaly distribution of the Earth’s surface calculated by 3D spherical triangulation of all land and sea measurements. All points of the earth are covered using this method.

Upgrading from HadSST3 to HadSST4 has led to a slight increase in average temperature anomalies due to instrumentation corrections. The average temperature anomaly in March was 0.96C up from 0.78C in February.  The first three months of 2022 show a small temperature  increase on the annual 2021 average temperature from 0.79C to  0.86C

The monthly data show the effects of La Nina and El Nino cycles.

Monthly average temperatures calculated by the spherical triangulation method

The temperature data can be downloaded.

monthly anomalies

annual anomalies


Posted in AGW, Climate Change, climate science, Hadley, NOAA | Leave a comment

Wind Lulls

The installed capacity of  UK Wind farms is currently 25GW.  The business secretary Kwasi Kwarteng proposes to at least double that number, but there is a basic problem which seemingly everyone overlooks – Wind Lulls. Sometimes high pressure sits over northern Europe for many days on end bringing still air with no wind.  All UK, German and French wind turbines are becalmed producing little if any power. Life and essential services has to continue so old coal stations are fired up and CC Gas  stations run at maximum output to meet peak demand.

Here are two recent examples:

  1. The 7 day lull from 16-22 December 2021

Comparison of Gas and Wind output during the 7 day Wind Lull last December.

We also see below how coal is still needed to balance power on the grid.

2.The recent  10 day wind lull lasting 10 days. At least this time there was a bit of sunshine yet notice again how the remaining 2 coal power stations were also needed . The whole of March saw only two brief spells of good wind output.

10 day wind lull March 20 – March 30

Most of March saw light winds. Output reached a maximum of 15GW briefly or a maximum  load capacity of  60% . Note that I am also correcting the metered wind output to include embedded  small wind farms as well.

Gas output compared to wind output – almost perfect anti-correlation.

Don’t worry though we are told. We just need “energy storage”, but no-one ever calculates just how much energy we would need store in order to see us through a wind drought like we have just experienced or the one last December.

In December there was additionally no solar power generated. In fact solar energy is perversely anti-correlated to demand. Annual peak demand is around 6pm on winter evenings when solar energy output is zero. So let’s estimate how much energy would need to be stored to cover the December lull.

We need 7 days of continuous power delivery at an average load of  30GW. So we need to store:

7 x 24 x 30 = 5040 GWh  or  1.8 x 10**16 joules

This is a huge amount of energy which is approximately equivalent to

  • 1200 Hiroshima size bombs
  • 373 million fully charged Tesla Powerwalls
  • 67.2 million long range Tesla 3 car battery charges

So it is unlikely that any future fleet of electric cars can back up the grid, assuming their owners would agree to walk rather than drive during a wind lull.

The largest energy store in the UK is the Dinorwig pumped storage Power Station in Snowdonia. It took 10 years to construct but actually paid for itself within 2 years by balancing peak time loads.  It can store up to 9.1 GWh of energy which is a useful power source over short periods.  However it is still  500 times too small to balance a wind dominated energy grid for a week. Nor do we have enough mountains to dramatically increase such pumped storage systems.

The largest Tesla grid size battery storage is in Hornsdale Southern Australia. It can store 193 MWh which is useful to cover short outages but still way too small for a wind lull.

As David MacKay used to say “We need an Energy Policy which adds up”.


Posted in Energy, renewables, wind farms | 77 Comments