A first look at HadCRUT5

The latest version of the CRU station data is called CRUTEM5 which when combined with the new Met Office sea surface temperature data HadSST4 becomes the new official global temperature dataset HadCRUT5. This was released about 2 weeks ago and is now operational. We have already seen that HadSST4  increases recent temperatures mainly by updating the historic corrections of bucket and engine room intake temperatures. In addition CRU have used  a method of infilling sparse 5 degree bins similar to that of Cowtan & Way, which now becomes obsolete. I decided to have a look at the new data.

The main difference for programmers is that all the data are now stored in Net CDF files, which means you have to rewrite all the data handling software. The particular  format they use is not documented as far as I can tell. So I used the NCDUMP utility to work out the structure of their  NetCDF files. All the station files are now also in netCDF files instead of simple text files.

My method of calculating  global temperatures uses a spherical triangulation of measurement points across the surface of the earth. This has the advantage of naturally covering the full surface area. The size of individual triangles just changes with measurement density. So the grid “cells” are 3D triangles rather than 2D Lat,Lon cells. The temperature of each triangle is the average of the 3 vertices. After a couple of days effort I now have it working.

HadCRUT5 temperature distribution for December 2020 determined by spherical triangulation

So how do my results compare with theirs? Here is a direct comparison for the annual temperature anomalies.

Comparison of the official HadCRUT5 annual temperature anomalies and the Spherical triangulation method.

My results are remarkably similar to theirs whereas this was not the case with HadCRUT4. The reason why they are now in good agreement is because HadCRUT5 extrapolates a fit into empty grid cells using a method similar to that of Cowtan & way. As a result HadCRUT5 has now become one of the “warmer” datasets whereas HadCRUT4 was one of the coolest. Despite this CRUTEM5 has not really expanded much in station data. I find 10631 stations but of these only 7734 have data within the normalisation period 1961-1990, which is not much more than in CRUTEM4 (10295 & 7680).

Finally here are the monthly results.

HadCRUT5 Monthly temperature anomalies. calculated using spherical triangulation.

My monthly and annual temperature anomalies can both be viewed as simple text files.

The software written in IDL is also available on request.

Posted in climate science, CRU, Hadley, UK Met Office | Tagged , | 3 Comments

2020 global temperature equals that of 2016

The global temperature for 2020 was 0.88C equaling that in 2016 based on GHCN-V4C (homogenised) and HadSST3. The V4U (uncorrected) result is 0.82C for 2020 making it slightly cooler than 2016. These are based on the Spherical Triangulation technique described here.

Global temperature anomalies relative to a baseline 1961-1990, calculated using spherical triangulation.

December saw a large drop in global temperatures relative to November mainly due to a strong La Nina developing in the South Pacific.

December 2020 shows a strong La Nina

The monthly time series looks as follows.

Monthly temperature time series showing significant differences between raw and homogenised data since November 2018. 0.61C is the long term average since 2003.

Nick Stokes (TempLS) uses a similar algorithm but with a loess interpolation. It is interesting to compare his results to mine. I think he is using only the uncorrected V4U data combined with ERSST sea surface temperatures.


Nov 2020 0.891 Ave 2016 0.857
Dec 2020 0.628 Ave 2020 0.852
Diff 0.263 Diff 0.005

My Uncorrected V4U/HadSST3

Nov 2020 0.941 Ave 2016 0.838
Dec 2020 0.645 Ave 2020 0.818
Diff 0.246 Diff 0.02

My Corrected V4C/HadSST3

Nov 2020 0.960 Ave 2016 0.880
Dec 2020 0.655 Ave 2020 0.887
Diff 0.305 Diff -0.003

There are some small systematic differences between the results, but we can draw the following conclusions.

  1. 2020 reached the same temperature as 2016.
  2. There was a significant drop in temperature in December due to a strong La Nina.
  3. Consequently 2021 should start cooler
Posted in Uncategorized | 4 Comments

Nights warm faster than days

The world has warmed on land by up to 1.5 C since pre-industrial times, but what does that really mean? Most people assume it means extreme temperatures  are increasing. However that is not the full story. The daily change in temperature from night to day we experience on  land is far higher than that from global warming. Scientists use temperature anomalies  to estimate climate change based on  whether any location has warmed relative to a standard 30 year normalisation period. All reference data (CRU, GHCN, BERKELEY) are based on the 12 monthly anomalies of  average temperatures (Tav) for each station,  where Tav = (Tmax+Tmin)/2 processed from the original station records.  The only source for these original raw weather station data that I know of  is GHCN Daily. This 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. These also measure the diurnal temperature range because Tmin occurs at night and Tmax occurs around midday.

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

Trange (Daily Diurnal Range) = Tmax – Tmin

Robert Rhode from Berkeley Earth was the first person AFAIK to calculate global the Land temperature anomaly based solely on this daily data. I then repeated the calculation using  an icosahedral gridding method. First here are the Berkeley Earth Results.

Daily temperature anomalies. Traditional BEST temperatures are shown in red

The reduction in the temperature range is clearly visible in the annual data. This becomes clearer if we plot annual avergaes

Tav and Trange anomalies (Berkeley Earth) Trends

It is clear the the diurnal temperature range has decreased by about 0.6C since 1880. My analysis of GHCN Daily is described here. The next two  plots shows my independently calculated  result.

GHCN-Daily monthly anomalies. In red are shown the TMAX-TMIN (Diurnal) anomalies.

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

Again we see that Trange (TMax – TMin) has reduced significantly since 1880. The early data actually seem to show an increase prior to 1850 but stations data are sparse in this early period. I also find a reduction in Trange of about 0.6C since 1900.

Further evidence that minimum temperatures are warming  faster than maximum temperatures can be seen in the Australia ACORN  daily station data  of 112 stations. For each station I take the highest temperature recorded each year and the minimum temperature recorded each year. I then average both of these over the area of Australia based on station locations.

Average Maximum and Minimum temperatures in Australia

So in general the extreme maximum temperatures have not increased significantly in Australia whereas the coldest temperature (at night in winter) have.

All these examples essentially imply that 20th century warming has been mostly  occurring at night. Nights are warming faster than days. I think this can be explained as follows.

During the day the sun heats the surface which then cools mainly by convection through the troposphere up to where it can radiate to space. At night the inverse is true. Without any solar heating and clear skies, the surface cools until convection slows or stops and surface radiative cooling becomes more important. The surface cools as the radiation balance changes.  That is also how fog and ground frost occur. So increasing CO2 levels reduce radiative transfer to space more at night than it does during the day, when convection is more important. The increase in  the effective emission height affects both maximum and minimum temperatures but proportionally more so at night than during the day. This is illustrated below

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

Nights  shifts the temperature gradient towards Pure radiative cooling.

Posted in climate science | Tagged | 11 Comments