135: Norway - temperature trends STABLE before 1980

Over the next four posts I will look at the temperature data of Scandinavia starting with Norway. Like the rest of Scandinavia, Norway has some of the best temperature data in Western Europe despite having large areas with low population densities and harsh Arctic climates. Some of the temperature data extends back to the mid-eighteenth century. What this data shows is that for two hundred years up to 1980 the climate was more or less stable and exhibited no significant temperature increase. Then around 1988 the mean temperature appears to jump abruptly by about 1°C. This pattern is also seen in much of the rest of Europe as I first demonstrated in Post 44.

In total Norway has nineteen long stations with over 1200 months of data before 2014 and 78 medium stations with over 480 months of data (for a full list of stations see here). Their approximate locations are shown on the map in Fig. 135.1 below. The stations are fairly evenly distributed across the country with thirty of the stations lying within the Arctic Circle, but there does appear to be a greater concentration in the south and a relatively low number of stations between Trondheim and Tromsø in the middle of the country. It is also apparent that over 60% of stations are located on or near the coast.

 

Fig. 135.1: The (approximate) locations of the 97 longest weather station records in Norway. Those stations with a high warming trend between 1911 and 2010 are marked in red while those with a cooling or stable trend are marked in blue. Those denoted with squares are long stations with over 1200 months of data, while diamonds denote medium stations with more than 480 months of data.

 

In order to quantify the changes to the climate of Norway the temperature anomalies for all stations with over 480 months of data before 2014 were determined and averaged. This was done using the usual method as outlined in Post 47 and involved first calculating the temperature anomaly each month for each station, and then averaging those anomalies to determine the mean temperature anomaly (MTA) for the country. This MTA is shown as a time series in Fig. 135.2 and clearly shows that temperatures were fairly stable up until 1980. However at some point in the 1980s (probably in 1988) the mean temperature appears to increase abruptly by about 1°C.

 

Fig. 135.2: The mean temperature change for Norway since 1880 relative to the 1961-1990 monthly averages. The best fit is applied to the monthly mean data from 1881 to 1980 and has a positive gradient of +0.23 ± 0.16 °C per century.

 

The process of determining the MTA in Fig. 135.2 involved first determining the monthly reference temperatures (MRTs) for each station using a set reference period, in this case from 1961 to 1990, and then subtracting the MRTs from the raw temperature data to deliver the anomalies. If a station had at least twelve valid temperatures per month within the MRT interval then its anomalies were included in the calculation of the mean temperature anomaly (MTA). The total number of stations included in the MTA in Fig. 135.2 each month is indicated in Fig. 135.3 below. The peak in the frequency between 1960 and 1990 suggests that the 1961-1990 interval was indeed the most appropriate to use for the MRTs.

 

Fig. 135.3: The number of station records included each month in the mean temperature anomaly (MTA) trend for Norway in Fig. 135.2.

 

The data in Fig. 135.3 indicates that the greatest coverage of the country for temperature data is after 1950 with up to 93 long and medium stations in operation at any one time. This drops to about twenty in 1930 and to less than five before 1850. This means that the MTA for Norway before 1890 will be less reliable than its values after 1950. Note that a reliable MTA generally needs data from at least sixteen stations (see Post 57 for evidence). However if we calculate the MTA for Norway back to 1760 we obtain the trends shown in Fig. 135.4 below. These show that the MTA remains stable for much of the two hundred years before 1980 but the noise level increases before 1800 when the MTA is dependent on only one station: Trondheim.

 

Fig. 135.4: The mean temperature change for Norway since 1760 relative to the 1961-1990 monthly averages. The best fit is applied to the monthly mean data from 1781 to 1980 and has a slight positive gradient of +0.06 ± 0.06 °C per century.

 

If we next consider the change in temperature based on Berkeley Earth (BE) adjusted data we get the MTA data in Fig. 135.5 below. This again was determined by averaging each monthly anomaly from the 97 longest stations and suggests that the climate was fairly stable before 1880 but then warmed by over 1°C thereafter. In fact the 10-year average suggests a warming of over 1.5°C. Not only that but the warming is more continuous in nature than the raw data in Fig. 135.4 actually shows.

 

Fig. 135.5: Temperature trends for Norway based on Berkeley Earth adjusted data. The best fit linear trend line (in red) is for the period 1876-2010 and has a positive gradient of +0.86 ± 0.04°C/century.

 

Comparing the curves in Fig. 135.5 with the published Berkeley Earth (BE) version for Norway in Fig. 135.6 below we see that there is good agreement between the two sets of data. This indicates that the simple averaging of anomalies used to generate the BE MTA in Fig. 135.5 using adjusted data is as effective and accurate as the more complex gridding method used by Berkeley Earth in Fig. 135.6. In which case simple averaging should be just as effective and accurate in generating the MTA using raw unadjusted data in Fig. 135.4 and Fig. 135.2. In other words, the discrepancy between the adjusted data in Fig. 135.5 and the unadjusted data in Fig. 135.4 cannot be due to the averaging process. Any form of weighted averaging would also not affect the results.

 

Fig. 135.6: The temperature trend for Norway since 1750 according to Berkeley Earth.

 

Most of the differences between the MTA in Fig. 135.4 and the BE versions using adjusted data in Fig. 135.6 are instead mainly due to the data processing procedures used by Berkeley Earth. These include homogenization, gridding, Kriging and most significantly breakpoint adjustments. These lead to changes to the original temperature data, the magnitude of these adjustments being the difference in the MTA values seen in Fig. 135.4 and Fig. 135.5.

The magnitudes of these adjustments are shown graphically in Fig. 135.7 below. The blue curve is the difference in MTA values between adjusted (Fig. 135.5) and unadjusted data (Fig. 135.4), while the orange curve is the contribution to those adjustments arising solely from breakpoint adjustments. The overall adjustment from 1880 to 2013 is small, less than +0.2°C. The main impact is to change the shape of the long term trend from a step-like jump in Fig. 135.4 to a more continuous increase in Fig. 135.5. This involves raising temperatures between 1920 and 1980 by 0.2°C while lowering slightly temperatures before 1920.

Fig. 135.7: The contribution of Berkeley Earth (BE) adjustments to the anomaly data in Fig. 135.5 after smoothing with a 12-month moving average. The blue curve represents the total BE adjustments including those from homogenization. The linear best fit (red line) to these adjustments for the period 1876-2010 has a positive gradient of +0.143 ± 0.005 °C per century. The orange curve shows the contribution just from breakpoint adjustments.

Summary

According to the raw unadjusted temperature data, the climate of Norway remained stable for over two hundred years up until the 1980s (see Fig. 135.4). Then it suddenly increased in temperature by 1°C. Why?

In contrast, adjusted temperature data from Berkeley Earth claims to show that the climate of Norway has warmed more or less continuously since 1860 by over 1.5°C (see Fig. 135.4 and Fig. 135.5).


Acronyms

BE = Berkeley Earth.

MRT = monthly reference temperature (see Post 47).

MTA = mean temperature anomaly.

List of all stations in Norway with links to their raw data files.

123: Svalbard - temperature trends STABLE to 2000

One of the most notable features of Svalbard is its location. It is one of the most northerly inhabited regions on the planet, located 900 km from the North Pole and almost 1000 km from the northern coast of Norway. As a result it is probably the best source of temperature data that we have for the Arctic, particularly given that there is no significant temperature data available within a radius of 800 km of the North Pole as I showed in Post 118.

That said, there are only seven  stations in Svalbard that have over 480 months of data. Of these only two have over 1000 months of data with one being a long station with over 1200 months of data. The remaining five are medium stations with over 480 months of data (for a list see here). Although the Svalbard archipelago is a part of Norway, its distance from Norway implies that its climate is likely to be rather different. For that reason I am studying it separately in this post rather than with Norway (which I will examine later).

Fig. 123.1: The mean temperature change for Svalbard since 1900 relative to the 1956-1985 monthly averages. The best fit is applied to the monthly mean data from 1921 to 2000 and has a negative gradient of -0.46 ± 0.37 °C per century.

 

What the data that we do have for Svalbard tells us is that the change in climate since 1920 is very similar to that seen for Jan Mayen in Post 122. The climate cooled slightly until 2000 and then warmed by about 2°C (see Fig. 123.1 above). However this warming is still smaller in magnitude than the variations seen in the 5-year average of up to 4°C.

In order to quantify the changes to the climate of Svalbard the temperature anomalies for each of the seven stations with the most data were determined and averaged. This was done using the usual method as outlined in Post 47 and involved first calculating the temperature anomaly each month for each station, and then averaging those anomalies to determine the mean temperature anomaly (MTA) for the region. The anomalies were determining relative to the monthly reference temperatures (MRTs) using a set reference period, in this case from 1956 to 1985. The total number of stations included in the MTA in Fig. 123.1 each month is indicated in Fig. 123.2 below. The peak in the frequency between 1970 and 1980 suggests that the 1956-1985 interval was indeed the most appropriate to use for the MRTs.

 

Fig. 123.2: The number of station records included each month in the mean temperature anomaly (MTA) trend for Svalbard in Fig. 123.1.

 

The locations of the seven stations with the most temperature data are shown in the map in Fig. 123.3 below. Most stations are clustered around Barentsburg with two situated on the outer islands of Hopen to the southeast and Bear Island (as in the Alistair Maclean novel and film) to the south.

 

Fig. 123.3: The (approximate) locations of the seven longest weather station records in Svalbard. Those stations with a high warming trend between 1911 and 2010 are marked in red while those with a cooling or stable trend are marked in blue. Those denoted with squares are long stations with over 1200 months of data, while diamonds denote medium stations with more than 480 months of data.

 

If we next consider the change in temperature based on Berkeley Earth (BE) adjusted data we get the MTA data shown in Fig. 123.4 below. This again was determined by averaging each monthly adjusted anomaly from the seven longest stations in Svalbard. The mean temperature follows a similar trajectory to that of the unadjusted data in Fig. 123.1 with temperatures over a 10-year average (orange curve) fluctuating by over 2°C and a large peak occurring around 1930. Temperatures in 2010 are also about 0.5°C higher than in 1930.

 

Fig. 123.4: Temperature trends for Svalbard based on Berkeley Earth adjusted data. The best fit linear trend line (in red) is for the period 1921-2000 and has a negative gradient of -0.72 ± 0.15°C/century.

 

Comparing the curves in Fig. 123.4 with the published Berkeley Earth (BE) version for Svalbard and Jan Mayen in Fig. 123.5 below shows that there is good agreement between the two sets of data even though Fig. 123.5 includes data from Jan Mayen. This indicates that the simple averaging of anomalies used to generate the BE MTA in Fig. 123.4 is as effective and accurate as the more complex gridding method used by Berkeley Earth in Fig. 123.5. In which case simple averaging should be just as effective and accurate in generating the MTA using raw unadjusted data in Fig. 123.1 even though the geographical distribution of stations is far from homogeneous, as was shown in Fig. 123.3. What is truly remarkable about the graph in Fig. 123.5 below is that it suggests that Berkeley Earth thinks it can determine the mean temperatures for Svalbard and Jan Mayen as far back as 1760 even though there is no data before 1910.

 

Fig. 123.5: The temperature trend for Svalbard and Jan Mayen since 1750 according to Berkeley Earth.

 

The similarity in the two sets of data, Fig. 123.1 and Fig. 123.4, is reflected in the scale of the adjustments made to the original data by Berkeley Earth. The magnitudes of these adjustments are shown graphically in Fig. 123.6 below. The blue curve is the difference in MTA values between adjusted (Fig. 123.4) and unadjusted data (Fig. 123.1), while the orange curve is the contribution to those adjustments arising solely from breakpoint adjustments.

In this case neither set of adjustments is particularly large. The most significant adjustment is for data after 1990 which is adjusted downwards apparently to reduce the extent of the temperature rise seen after 1990 in Fig. 123.1. The other significant adjustment is for data between 1940 and 1960. The effect of this appears to be to reduce the size of the temperature peak before 1960. Neither of these adjustments significantly affect the overall trend of the data other than to reduce the warming after 2000 to values similar to those seen in BE adjusted trends for other regions.

 

Fig. 123.6: The contribution of Berkeley Earth (BE) adjustments to the anomaly data in Fig. 123.4 after smoothing with a 12-month moving average. The blue curve represents the total BE adjustments including those from homogenization. The linear best fit (red line) to these adjustments for the period 1921-2000 has a slight negative gradient of -0.13 ± 0.03 °C per century. The orange curve shows the contribution just from breakpoint adjustments.

 

Summary

According to the raw unadjusted temperature data, the climate of Svalbard has remained fairly stable since 1930 with possibly some cooling, but may have warmed by up to 2°C since 2000 (see Fig. 123.1). Any warming since 2000 is still comparable to the natural variations in temperature seen over the previous eighty years.

Over the same period the adjusted temperature data from Berkeley Earth appears to show similar climatic variations (see Fig. 123.4) to the unadjusted (see Fig. 123.1).

Berkeley Earth has estimated the climate variations for Svalbard as far back as 1760 (see Fig. 123.5) despite there being no reliable temperature data before 1910.

The patterns seen in the temperature data for Svalbard in Fig. 123.1 (decline before 1990 and warming after) are similar to those seen previously for nearby islands of Greenland (Fig. 119.1 in Post 119), Iceland (Fig. 120.1 in Post 120), the Faroes (Fig. 121.1 in Post 121) and Jan Mayen (Fig. 122.2 in Post 122).



Acronyms

BE = Berkeley Earth.

MRT = monthly reference temperature (see Post 47).

MTA = mean temperature anomaly.

Link to list of all stations in Svalbard and their raw data files.

122: Jan Mayen - temperature trends VARIABLE

The Norwegian island of Jan Mayen lies in the Arctic Circle approximately 600 km northeast of Iceland, 500 km east of Greenland and almost 1000 km from the coast of Norway. It is just over 50 km long and is inhabited only by a few Norwegian military and meteorological personnel on six month deployments. It does, though, have two weather stations (BE ID: 16234 and BE ID: 157312 with 735 and 1113 months of data respectively) that provide the only temperature data for over 500 km in all directions (see Fig. 122.1 below).

 

Fig. 122.1: The (approximate) locations of the two longest weather station records in Jan Mayen. Those stations with a high warming trend between 1911 and 2010 are marked in red while those with a cooling or stable trend are marked in blue.

 

The mean temperature anomaly (MTA) for Jan Mayen is shown in Fig. 122.1 below. From 1920 to 1990 the temperature trend is downward with temperatures falling by about 0.7°C according to the best fit line (red line) or 1.5°C according to the 5-year moving average (yellow line). After 1990 the temperature increases by up to 2°C. Both of these changes are less than the natural variation seen in the 5-year average (yellow curve) so it is impossible to definitively attribute them to climate change.

 

Fig. 122.2: The mean temperature change for Jan Mayen since 1920 relative to the 1971-2000 monthly averages. The best fit is applied to the monthly mean data from 1921 to 2000 and has a negative gradient of -0.93 ± 0.25 °C per century.

The MTA in Fig. 122.2 is the average of anomaly data from two stations (see Fig. 122.3 below) and was determined using the usual method as outlined in Post 47. The anomalies were determined relative to monthly reference temperatures (MRT) with the MRTs calculated using data from 1971-2000.

Fig. 122.3: The number of station records included each month in the mean temperature anomaly (MTA) trend for Jan Mayen in Fig. 122.2.

 

Repeating the averaging process using data that has been adjusted by Berkeley Earth (BE) yields the temperature curve shown in Fig. 122.4 below. In this case the 10-year average (orange curve) is broadly similar in shape to the yellow curve in Fig. 122.2, but the temperature fall before 1990 and the rise after are both slightly larger. In both cases, however, the temperatures in 2010 are less than 0.5°C higher than in 1930.

 

Fig. 122.4: Temperature trends for Jan Mayen based on Berkeley Earth adjusted data. The best fit linear trend line (in red) is for the period 1921-2000 and has a negative gradient of -1.04 ± 0.11°C/century.

 

The similarity in the two sets of data (Fig. 122.2 and Fig. 122.4) is reflected in the scale of the adjustments made to the original data by Berkeley Earth. The magnitudes of these adjustments are shown graphically in Fig. 122.5 below. The blue curve is the difference in MTA values between adjusted (Fig. 122.4) and unadjusted data (Fig. 122.2), while the orange curve is the contribution to those adjustments arising solely from breakpoint adjustments. In this case neither are particularly large, with data after 1950 being adjusted down by about 0.18°C. The only significant adjustment is for data between 1960 and 1976 which is adjusted upwards by about 0.9°C. The effect of this is to reduce the size of the temperature dip before 1970 seen in Fig. 122.2.

 

Fig. 122.5: The contribution of Berkeley Earth (BE) adjustments to the anomaly data in Fig. 122.4 after smoothing with a 12-month moving average. The blue curve represents the total BE adjustments including those from homogenization. The linear best fit (red line) to these adjustments for the period 1921-2000 has a slight negative gradient of -0.07 ± 0.04 °C per century. The orange curve shows the contribution just from breakpoint adjustments.

 

Summary

According to the raw unadjusted temperature data, the climate of Jan Mayen has remained fairly stable since 1930 (see Fig. 122.2).

Over the same period adjusted temperature data from Berkeley Earth appears to show similar climate variations (see Fig. 122.5).

The patterns seen in the temperature data for Jan Mayen in Fig. 122.2 (decline before 1990 and warming after) are similar to those seen previously for nearby islands of Greenland (Fig. 119.1 in Post 119), Iceland (Fig. 120.1 in Post 120) and the Faroes (Fig. 121.1 in Post 121).



Acronyms

BE = Berkeley Earth.

MRT = monthly reference temperature (see Post 47).

MTA = mean temperature anomaly.

121: Faroe Islands - temperature trends STABLE

The longest temperature record in the Faroe Islands is from a station at Thorshavn. It has nearly 1800 months of data, but no other temperature record in the islands has more than 435 months of data. In fact only another two (at Vagar and Akraberg) have more than 120 months of data (see here for a list). This means that the temperature record for the Faroe Islands is heavily dependent on the Thorshavn data with the other two stations only affecting the mean temperatures after 1980. 

The result is the time series for the mean temperature anomaly (MTA) shown in Fig. 121.1 below. What is striking is that the general form of the five-year moving average (yellow curve) is very similar to those for both Greenland (see Fig. 119.1 in Post 119) and Iceland (see Post 120.1 in Post 120). Temperatures declined from 1930 until 1990 and then rebounded. In all three cases the temperatures today are no higher than in the 1930s but the amount of cooling from 1930-1990 is different, being highest for Greenland and least for the Faroe Islands. This is not a surprise, though, as it is well known that climates that are more extreme (i.e. Greenland) experience larger temperature fluctuations than are seen in more temperate climes (i.e. the Faroe Islands).

Fig. 121.1: The mean temperature change for the Faroe Islands since 1920 relative to the 1981-2010 monthly averages. The best fit is applied to the monthly mean data from 1931 to 2010 and has a negative gradient of -0.40 ± 0.15 °C per century.

In order to quantify the changes to the climate of the Faroe Islands the temperature anomalies for the three longest station records were averaged. This was done using the usual method as outlined in Post 47 and involved first calculating the temperature anomaly each month for each station, and then averaging those anomalies to determine the mean temperature anomaly (MTA) for the region.

The process of determining the MTA in Fig. 121.1 involved first determining the monthly reference temperatures (MRTs) for each station using a set reference period, in this case from 1981 to 2010, and then subtracting the MRTs from the raw temperature data to deliver the anomalies. The total number of stations included in the MTA in Fig. 121.1 each month is indicated in Fig. 121.2 below. The peak in the frequency after 1980 illustrates why the 1981-2010 interval was indeed the most appropriate to use for the MRTs.

Fig. 121.2: The number of station records included each month in the mean temperature anomaly (MTA) trend for the Faroe Islands in Fig. 121.1.

If all the data is considered, the MTA trend for the Faroe Islands will have data extending back to 1867 as shown in Fig. 121.3 below. Much of the data before 1930 indicates that temperatures then, at least for Thorshavn, were cooler than in the 1930s and to day. Again this is similar to the results from Greenland and Iceland, but it is unclear how much of this temperature change is genuine warming of the climate and how much is just the result of natural fluctuations. But the fact that these features are repeated in multiple regions means that they cannot be discounted as being isolated results that are the result of poor measurements.

Fig. 121.3: The mean temperature change for the Faroe Islands since 1860 relative to the 1981-2010 monthly averages. The best fit is applied to the monthly mean data from 1871 to 2010 and has a positive gradient of +0.27 ± 0.07 °C per century.

The locations of the three stations used to determine the MTA in Fig. 121.3 are shown in the map in Fig. 11214 below. The temperature data from all three appear to exhibit modest warming trends, but this is mainly due to the fitting intervals used in each case. Vagar and Akraberg only have data after 1970 where the MTA in Fig. 121.3 is warming. The trend for Thorshavn was for data from 1911-2010, and is therefore biased by the fact that the fitting interval does not extend from peak to peak. This is akin to fitting to a sine wave as I explained with Fig. 4.7 in Post 4. So the positive trend is not real.

Fig. 121.4: The (approximate) locations of the three longest weather station records in the Faroe Islands. Those stations with a high warming trend between 1911 and 2010 are marked in red while those with a cooling or stable trend are marked in blue. Those denoted with squares are long stations with over 1200 months of data, while diamonds denote stations with more than 240 months of data.

If we next consider the change in temperature based on Berkeley Earth (BE) adjusted data we get the MTA data shown in Fig. 121.5 below. This again was determined by averaging each monthly anomaly from the three longest stations in the Faroe Islands. The mean temperature follows a similar trajectory to that of the unadjusted data in Fig. 121.3 with temperatures fluctuating by over 1°C and a large peak occurring around 1930. However the BE adjustments appear to have lowered this peak relative to temperatures in 2010 by over 0.25°C compared to the raw data in Fig. 121.3.

Fig. 121.5: Temperature trends for the Faroe Islands based on Berkeley Earth adjusted data. The best fit linear trend line (in red) is for the period 1871-2010 and has a positive gradient of +0.61 ± 0.03°C/century.

Comparing the curves in Fig. 121.5 with the published Berkeley Earth (BE) version for the Faroe Islands in Fig. 121.6 below shows that there is good agreement between the two sets of data. However, Berkeley Earth appear to think they can determine the temperature back to 1760 even though there is no data before 1860. Who says climate scientists are pessimists?

Fig. 121.6: The temperature trend for the Faroe Islands since 1750 according to Berkeley Earth.

The differences between the MTA in Fig. 121.3 and the BE versions using adjusted data in Fig. 121.5  can be determined by calculating the difference in the MTA values seen in Fig. 121.3 and Fig. 121.5. The result is shown graphically in Fig. 121.7 below. The blue curve is the difference in MTA values between adjusted (Fig. 121.5) and unadjusted data (Fig. 121.3), while the orange curve is the contribution to those adjustments arising solely from breakpoint adjustments. What is clear is that there is only one significant adjustment. That is a breakpoint adjustment of over 0.25°C made to the Thorshavn data in 1951. The difference between the blue and orange curves in Fig. 121.7 is due to the difference in MRT interval used in Fig. 121.3 and by Berkeley Earth in Fig. 121.5.

Fig. 121.7: The contribution of Berkeley Earth (BE) adjustments to the anomaly data in Fig. 121.5 after smoothing with a 12-month moving average. The blue curve represents the total BE adjustments including those from homogenization. The linear best fit (red line) to these adjustments for the period 1871-2010 has a positive gradient of +0.330 ± 0.005 °C per century. The orange curve shows the contribution just from breakpoint adjustments.

Summary

According to the raw unadjusted temperature data, the climate of the Faroe Islands has cooled from 1930 to 1980 by about 1°C. It then warmed by a similar but slightly smaller amount until 2005 (see Fig. 121.1).

Over the same period adjusted temperature data from Berkeley Earth appears to show that the climate of the Faroe Islands has warmed by over 0.25°C since 1930 and over 1°C since the 1800s (see Fig. 121.5).

The difference in the raw unadjusted data (Fig. 121.3) and the adjusted data (Fig. 121.5) is mainly due to a single breakpoint adjustment of 0.26°C in 1951.

Any warming trend since 1870 is small (~0.3°C) compared to what look like natural temperature variations (~1°C). The origin of these variations (shown in Fig. 121.3) is uncertain but cannot be the result of greenhouse gas emissions when those emissions were so low compared to today. However, similar patterns are seen in the temperature data of nearby islands of Greenland (Post 119), Iceland (Post 120). and Jan Mayen (from 1920 only), so these features seen in the data before 1930 may be real and representative of synchronous climate variations occurring across the North Atlantic/Arctic region.



Acronyms

BE = Berkeley Earth.

MRT = monthly reference temperature (see Post 47).

MTA = mean temperature anomaly.

Link to list of all stations in the Faroe Islands and their raw data files.

120: Iceland - temperature trends COOLING before 2000

It is tempting to think of Iceland as being Greenland's little brother. This is not just because Iceland is smaller and close to Greenland, but also because their changes in climate over the last two hundred years are very similar. Like in Greenland, the climate of Iceland cooled from 1930 to 1990 before the mean temperatures rebounded. And like in Greenland, the mean temperatures today are no higher than they were in the 1930s (see Fig. 120.1 below).

Fig. 120.1: The mean temperature change for Iceland since 1920 relative to the 1971-2000 monthly averages. The best fit is applied to the monthly mean data from 1931 to 2000 and has a negative gradient of -1.43 ± 0.24 °C per century.

In order to quantify the changes to the climate of Iceland the temperature anomalies for each of the 21 stations with the most data (over 300 months) were determined and averaged. This was done using the usual method as outlined in Post 47 and involved first calculating the temperature anomaly each month for each station, and then averaging those anomalies to determine the mean temperature anomaly (MTA) for the region. The MTA since 1920 is shown as a time series in Fig. 120.1 above and clearly shows that temperatures declined continuously from 1930 to 1990 before levelling off and then rebounding.

The process of determining the MTA in Fig. 120.1 involved first determining the monthly reference temperatures (MRTs) for each station using a set reference period, in this case from 1971 to 2000, and then subtracting the MRTs from the raw temperature data to deliver the anomalies. If a station had at least twelve valid temperatures per month within the MRT interval (1971-2000) then its anomalies were included in the calculation of the mean temperature anomaly (MTA). The total number of stations included in the MTA in Fig. 120.1 each month is indicated in Fig. 120.2 below. The peak in the frequency after 1980 suggests that the 1971-2000 interval was indeed the most appropriate to use for the MRTs, although 1981-2010 would have been equally appropriate.

Fig. 120.2: The number of station records included each month in the mean temperature anomaly (MTA) trend for Iceland in Fig. 120.1.

The data in Fig. 120.2 above indicates that after 1940 there were up to 21 active stations, but before 1940 there were less than about six with only one station being operational before 1870. As six is generally too low a number to produce a reliable trend, the MTA data in Fig. 120.1 was truncated with only data post-1920 being shown. However, if all the data is considered, the MTA trend will have data extending back to 1823 as shown in Fig. 120.3 below. Note also that the low number of stations before 1900 results in a much higher variance of points in Fig. 120.3 about the mean (yellow line). This is more evidence of the greater unreliability of this earlier data, which is why the plot shown in Fig. 120.1 is more statistically reliable.

Fig. 120.3: The mean temperature change for Iceland since 1820 relative to the 1971-2000 monthly averages. The best fit is applied to the monthly mean data from 1841 to 2000 and has a positive gradient of +0.71 ± 0.08 °C per century.

The locations of the 21 stations used to determine the MTA in Fig. 120.3 are shown in the map in Fig. 120.4 below. Of these 21 stations, six are long stations with over 1200 months of data before 2014, and a further five are medium stations with over 480 months of data. The stations are evenly distributed across the island with most on, or near, the coast.

Fig. 120.4: The (approximate) locations of the 21 longest weather station records in Iceland. Those stations with a high warming trend between 1911 and 2010 are marked in red while those with a cooling or stable trend are marked in blue. Those denoted with squares are long stations with over 1200 months of data, while diamonds denote stations with more than 300 months of data.

If we next consider the change in temperature based on Berkeley Earth (BE) adjusted data we get the MTA data shown in Fig. 120.5 below. This again was determined by averaging each monthly anomaly from the 21 longest stations in Iceland. The mean temperature follows a similar trajectory to that of the unadjusted data in Fig. 120.3 with temperatures fluctuating by over 1°C and a large peak occurring around 1930. However the BE adjustments appear to have lowered this peak slightly relative to temperatures in 2010 when compared to the raw data in Fig. 120.3.

Fig. 120.5: Temperature trends for Iceland based on Berkeley Earth adjusted data. The best fit linear trend line (in red) is for the period 1841-2010 and has a positive gradient of +0.51 ± 0.03°C/century.

 

Comparing the curves in Fig. 120.5 with the published Berkeley Earth (BE) version for Iceland in Fig. 120.6 below shows that there is good agreement between the two sets of data. This indicates that the simple averaging of anomalies used to generate the BE MTA in Fig. 120.5 is as effective and accurate as the more complex gridding method used by Berkeley Earth in Fig. 120.6. In which case simple averaging should be just as effective and accurate in generating the MTA using raw unadjusted data in Fig. 120.1.

Fig. 120.6: The temperature trend for Iceland since 1750 according to Berkeley Earth.

Most of the differences between the MTA in Fig. 120.3 and the BE versions using adjusted data in Fig. 120.6  are mainly due to the data processing procedures used by Berkeley Earth. These include homogenization, gridding, Kriging and most significantly breakpoint adjustments. These lead to changes to the original temperature data, the magnitude of these adjustments being the difference in the MTA values seen in Fig. 120.3 and Fig. 120.5.

The magnitudes of these adjustments are shown graphically in Fig. 120.7 below. The blue curve is the difference in MTA values between adjusted (Fig. 120.5) and unadjusted data (Fig. 120.1), while the orange curve is the contribution to those adjustments arising solely from breakpoint adjustments. Both result in a consistent upward trend after 1920 with the former leading to an additional warming since 1930 of up to 0.25°C. These adjustments are, however, much smaller in total than the natural variation seen in the raw data in Fig. 120.3, so while they change the overall magnitude of the climate changes slightly, the general form of the temperature trends in Fig. 120.5 and Fig. 120.3 look broadly similar.

Fig. 120.7: The contribution of Berkeley Earth (BE) adjustments to the anomaly data in Fig. 120.5 after smoothing with a 12-month moving average. The blue curve represents the total BE adjustments including those from homogenization. The linear best fit (red line) to these adjustments for the period 1921-2010 has a positive gradient of +0.234 ± 0.007 °C per century. The orange curve shows the contribution just from breakpoint adjustments.

Summary

According to the raw unadjusted temperature data, the climate of Iceland has cooled from 1930 to 1990 by about 1°C. It then warmed by a similar but slightly smaller amount until 2005 (see Fig. 120.1).

Over the same period adjusted temperature data from Berkeley Earth appears to show that the climate of Iceland has warmed only fractionally since 1930, but by up to 2°C since the 1800s (see Fig. 120.5).

The reliability of the temperature data before 1930 is debatable due to the low number of stations and the large jumps in temperature that occur repeatedly. The origin of these jumps is uncertain but cannot solely be the result of greenhouse gas emissions when those emissions increased the atmospheric carbon dioxide concentration by so little compared to today.


Acronyms

BE = Berkeley Earth.

MRT = monthly reference temperature (see Post 47).

MTA = mean temperature anomaly.

Link to list of all stations in Iceland and their raw data files.

113: The Guianas - temperature trends STABLE

The Guianas is the region of north-eastern South America that comprises the three territories of Guyana (formerly British Guiana), Suriname (formerly Dutch Guiana) and French Guiana. It sits on the Atlantic coast between Venezuela and Brazil, and as the data in Fig. 113.1 below shows, it does not appear to have experienced any significant climate change over the last 100 years, although the mean temperature has fluctuated significantly by up to 1°C (see yellow curve).

Fig. 113.1: The mean temperature change for the Guianas relative to the 1961-1990 monthly averages. The best fit is applied to the monthly mean data from 1896 to 2005 and has a slight positive gradient of +0.11 ± 0.04 °C per century.

The MTA in Fig. 113.1 was calculated by averaging the temperature anomalies from the fourteen longest temperature records for the region. Eight of these records were from medium stations with over 480 months of temperature data before the end of 2013, but there are only two long stations with more than 1200 months of data.

The anomalies for each station were determined using the usual method as outlined in Post 47. This involved first calculating the monthly reference temperatures (MRTs) for each station using a set reference period, in this case from 1961 to 1990, and then subtracting the MRTs from the raw temperature data to deliver the anomalies. If a station had at least twelve valid temperatures per month within the MRT interval then its anomalies were included in the MTA calculation. The total number of stations included in the MTA in Fig. 113.1 each month is indicated in Fig. 113.2 below. The peak just around 1975 suggests that the 1961-1990 interval was indeed the most appropriate.

Fig. 113.2: The number of station records included each month in the mean temperature anomaly (MTA) trend for the Guianas in Fig. 113.1.

The locations of the fourteen main stations are shown in the map in Fig. 113.3 below. This appears to show that the geographical spread is fairly uniform, although there does appear to be far more stations in Suriname than in either Guyana or French Guiana. This variation in station density is probably not sufficient to significantly distort the average in Fig. 113.1 from its true value though. In which case the simple average of the anomalies from all stations used to construct the MTA in Fig. 113.1 should still yield a fairly accurate temperature trend for the region as a whole. 

Overall there are more stations close to the coast than inland, and the coastal stations appear more likely to have warming trends. A warming trend is defined here as one where the temperature gradient for 1911-2010 is positive and exceeds twice the uncertainty in that trend.

Fig. 113.3: The (approximate) locations of the fourteen longest weather station records in the Guianas. Those stations with a high warming trend between 1911 and 2010 are marked in red while those with a cooling or stable trend are marked in blue. Those denoted with squares are long stations with over 1200 months of data, while diamonds denote stations with more than 300 months of data.

In contrast to Fig. 113.1, the corresponding MTA dataset based on data that has been adjusted by Berkeley Earth (BE) exhibits a strong warming trend with temperatures rising by over 1.5°C since 1890 (see Fig. 113.4 below).

Fig. 113.4: Temperature trends for the Guianas based on Berkeley Earth adjusted data. The best fit linear trend line (in red) is for the period 1896-2005 and has a gradient of +1.06 ± 0.03°C/century.

If we next compare the curves in Fig. 113.4 with the published Berkeley Earth (BE) version for Suriname in Fig. 113.5 below (where most stations are located) we see that there is remarkably good agreement between the two sets of data at least as far back as 1900. This indicates that the simple averaging of anomalies used to generate the BE MTA in Fig. 113.4 is as effective and accurate as the more complex gridding method used by Berkeley Earth in Fig. 113.5. In which case simple averaging should be just as effective and accurate in generating the MTA using raw unadjusted data in Fig. 113.1.

Fig. 113.5: The temperature trend for Suriname since 1820 according to Berkeley Earth.

The differences between the MTA in Fig. 113.1 and the BE versions using adjusted data in Fig. 113.4  are instead mainly due to the data processing procedures used by Berkeley Earth. These include homogenization, gridding, Kriging and most significantly breakpoint adjustments. These lead to changes to the original temperature data, the magnitude of these adjustments being the difference in the MTA values seen in Fig. 113.1 and Fig. 113.4. The magnitudes of these adjustments are shown graphically in Fig. 113.6 below. The blue curve is the difference in MTA values between adjusted (Fig. 113.4) and unadjusted data (Fig. 113.1), while the orange curve is the contribution to those adjustments arising solely from breakpoint adjustments. Both are considerable with the former leading to an additional warming since 1900 of up to 1.6°C.

Fig. 113.6: The contribution of Berkeley Earth (BE) adjustments to the anomaly data in Fig. 113.4 after smoothing with a 12-month moving average. The blue curve represents the total BE adjustments including those from homogenization. The linear best fit (red line) to these adjustments for the period 1896-2005 has a positive gradient of +0.96 ± 0.03 °C per century. The orange curve shows the contribution just from breakpoint adjustments.

Summary

According to the raw unadjusted temperature data, over the past century the climate of the Guianas has remained stable (see Fig. 113.1).

Over the same period adjusted temperature data from Berkeley Earth appears to show that the climate of the Guianas has warmed by as much as 1.6°C (see Fig. 113.4).


Acronyms

BE = Berkeley Earth.

MRT = monthly reference temperature (see Post 47).

MTA = mean temperature anomaly.

Link to list of all stations in Guyana and their raw data files.

Link to list of all stations in Suriname and their raw data files.

Link to list of all stations in French Guiana and their raw data files.

106. Bahamas and Key West - temperature trends STABLE to 1988

The islands of The Bahamas stretch over a distance of more than 800 km on the edge of the Atlantic Ocean southeast of Florida. Yet only four weather stations in the region have sufficient temperature data to be useful (for a list see here), and two of these are in Nassau (see map in Fig. 106.1 below). In addition, however, there are two stations in Key West (Key West and Key West airport) that are so far from the Florida coast as to be possibly more representative of the climate of The Bahamas than that of Florida (see Post 103). For this reason I will include them in this analysis.

Fig. 106.1: The (approximate) locations of the six longest weather station records in The Bahamas and Key West. Those stations with a high warming trend between 1911 and 2010 are marked in red while those with a cooling or stable trend are marked in blue. Those denoted with squares are long stations with over 1200 months of data, while diamonds denote medium stations with more than 480 months of data.

In total there are two long stations with over 1200 months of data before 2014 and four medium stations with over 480 months in this analysis. The two long stations both had more or less continuous data that stretched from before 1900 to 2013. The four medium stations were more problematic. Three had virtually no data before 1950 while the station at Nassau had no data after. Added to that, the station at Freeport airport had no data before 1970.

The temperature anomalies for each station were determined using the usual method as outlined in Post 47. This involved first calculating the monthly reference temperatures (MRTs) for each station using a set reference period, in this case from 1961 to 1990, and then subtracting the MRTs from the raw temperature data to deliver the anomalies. If a station had at least twelve valid temperatures per month within the MRT interval then its anomalies were included in the calculation of the regional mean temperature anomaly (MTA). 

As no one single time interval for the monthly reference temperatures (MRTs) would allow all six stations to be included in the final average, the MRT interval was set to be 1961-1990. The one station to be excluded from the MTA calculation in this case was the Nassau station, but this exhibits virtually zero temperature change over its data range from 1900 to 1950.

Fig. 106.2: The mean temperature change for The Bahamas and Key West relative to the 1961-1990 monthly averages. The best fit is applied to the monthly mean data from 1901 to 1980 and has a slight positive gradient of +0.11 ± 0.12 °C per century.

The resulting MTA is shown in Fig. 106.2 above. It can be seen that before 1988 there is only a very slight upward temperature trend that is less than the uncertainty in the trend. Then in 1988 the temperature jumps suddenly by about 0.5°C. This is similar to the jump of about 1°C that was identified earlier in Post 44 for the MTA of Europe which also occurred in or around 1988. Is this coincidence, or did something happen to data collection methods in 1988?

The total number of stations included in the MTA in Fig. 106.2 each month is indicated in Fig. 106.3 below. The peak in the frequency around 1980 suggests that the 1961-1990 interval was indeed the most appropriate, but it also shows how much of the MTA trend in Fig. 106.2 relies on data from just two stations: Nassau airport and Key West airport.

Fig. 106.3: The number of station records included each month in the mean temperature anomaly (MTA) trend for The Bahamas and Key West in Fig. 106.2.

Next I calculate the corresponding MTA result based on data that has been adjusted by Berkeley Earth (BE). The result is shown in Fig. 106.4 below.

Fig. 106.4: Temperature trends for The Bahamas and Key West based on Berkeley Earth adjusted data. The best fit linear trend line (in red) is for the period 1911-2010 and has a gradient of +0.79 ± 0.03°C/century.

 

Comparing the curves in Fig. 106.4 with the published Berkeley Earth (BE) version in Fig. 106.5 below indicates remarkably good agreement at least as far back as 1900 despite Fig. 103.4 also including data from Key West. This suggests that the simple averaging of anomalies I have used is effective and accurate, and adding the Key West stations was probably appropriate.

Fig. 106.5: The temperature trend for The Bahamas since 1750 according to Berkeley Earth.

The differences between the MTA in Fig. 106.2 and the BE versions using adjusted data in Fig. 106.4 and Fig. 106.5 are mainly due to the data processing procedures used by Berkeley Earth. These include homogenization, gridding, Kriging and most significantly breakpoint adjustments. These lead to changes to the original temperature data, the magnitude of these adjustments being the difference in the MTA values seen in Fig. 106.2 and Fig. 106.4. The magnitudes of these adjustments are shown graphically in Fig. 106.6 below. The blue curve is the difference in MTA values between adjusted (Fig. 106.4) and unadjusted data (Fig. 106.2), while the orange curve is the contribution to those adjustments arising solely from breakpoint adjustments. Both are considerable and produce an additional warming since 1900 of about 0.4°C.

Fig. 106.6: The contribution of Berkeley Earth (BE) adjustments to the anomaly data in Fig. 106.4 after smoothing with a 12-month moving average. The blue curve represents the total BE adjustments including those from homogenization. The linear best fit (red line) to these adjustments for the period 1901-1980 has a positive gradient of +0.682 ± 0.018 °C per century. The orange curve shows the contribution just from breakpoint adjustments.

 

Summary 

According to the raw unadjusted temperature data, over the ninety year period up to 1988 the climate of The Bahamas and Key West remained stable before experiencing a sudden jump in temperature of about 0.5°C (see Fig. 106.2).

Over the period 1901-2010 the adjusted temperature data from Berkeley Earth claims to show that the climate of The Bahamas and Key West has warmed by as much as 1.0°C (see Fig. 106.4).

Acronyms 

BE = Berkeley Earth.

MRT = monthly reference temperature (see Post 47).

MTA = mean temperature anomaly.

 

List of all stations

Nassau airport

Nassau

Abrahams Bay

Freeport airport

Key West

Key West airport

104. US southern states - summary of temperature trends

Over the last month I have examined the temperature trends of five different US states (Louisiana, Mississippi, Alabama, Georgia and Florida) that surround, or are within 100km of (in the case of Georgia), the Gulf of Mexico. These all appear to have similar trends to that of Texas that I examined in Post 52. All have negative or stable temperature trends over the last 100 years. For comparison their temperature trends are republished here with identical data ranges (from 1900) and fitting ranges (1911-2010). What is clear is that none of these trends is remotely similar to either the Berkeley Earth (BE) versions for each state based on adjusted data, or the global trends published by NOAA, NASA-GISS, BE, HadCRU etc.

Fig. 104.1: The mean temperature change for Texas. The best fit has a slight negative gradient of -0.15 ± 0.15 °C per century.

Fig. 104.2: The mean temperature change for Louisiana. The best fit has a negative gradient of -0.38 ± 0.15 °C per century.

Fig. 104.3: The mean temperature change for Mississippi. The best fit has a negative gradient of -0.76 ± 0.17 °C per century.

Fig. 104.4: The mean temperature change for Alabama. The best fit has a negative gradient of -0.72 ± 0.17 °C per century.

Fig. 104.5: The mean temperature change for Georgia. The best fit has a negative gradient of -0.76 ± 0.16 °C per century.

Fig. 104.6: The mean temperature change for Texas. The best fit has a slight positive gradient of +0.08 ± 0.13 °C per century.