What is striking about the temperature data for Peru is its variability. Not only is there a diverse mix of warming and cooling trends between stations, there are also a lot of extreme fluctuations within individual temperature records as illustrated by the time series for Arequipa Airport (Berkeley Earth ID:157461) in Fig. 63.1 below. This makes it very hard to assess what the true temperature trend for Peru really is.
Fig. 63.1: The temperature trend for Arequipa Airport since 1900. The best fit line has a positive gradient of +0.16 ± 0.14 °C per century. The monthly temperature changes are defined relative to the 1951-1980 monthly averages.
In all there are 42 stations in Peru with more than 300 months of data. Their locations are shown in Fig. 63.2 below. It can be seen that they are spread throughout most of Peru, but there is significant clustering in some regions and sparse coverage in others, particularly within the Amazon region to the north and east. Of these 42 stations, 24 are medium stations with over 480 months of data, the longest of which is Arequipa Airport (Berkeley Earth ID:157461) with 1163 months of data. There are no long stations with more than 1200 months of data.
Fig. 63.2: The (approximate) locations of all stations in Chile with over 300 months of data. Those stations with a high warming trend are marked in red. Those with cooling or stable trends are marked in blue.
The station location map in Fig. 63.2 indicates that there is a fairly even mix of warming and cooling stations in Peru. However, stations with data before 1960 are more likely to exhibit cooling trends, while those with data after 1960 are more likely to be warming. The result is that the overall temperature trend from 1930 onwards comprises a sharp cooling period followed by a slow warming as shown in Fig. 63.3 below.
Fig. 63.3: The temperature trend for Peru since 1900. The best fit is applied to all the monthly mean data and has a positive gradient of +1.11 ± 0.06 °C per century. The monthly temperature changes are defined relative to the 1951-1980 monthly averages.
The temperature trend in Fig. 63.3 above was derived by averaging the temperature anomalies from all the stations with more than 300 months of data which also had at least ten years of data within the interval of 1951-1980. This amounted to 39 stations in total (for a list see here). The interval of 1951-1980 was used to determine the monthly reference temperatures (MRTs) against which the temperature anomalies were determined, as explained in Post 47. This period was chosen so as to maximize the number of stations included in the final mean trend.
If we perform a fit to all the data in Fig. 63.3, the result is a strong warming trend of 1.11°C per century as indicated in Fig. 63.3 above. Not only does this appear to closely follow the data from 1960 onwards, it also appears to fit with the data before 1925 as well.
However, the data before 1925 comes from at most two stations, as indicated in Fig. 63.4 below, while the data from 1930 to 1960 in Fig. 63.3 is the result of averaging at least fifteen different temperature records from different stations, and potentially as many as thirty. This suggests that the mean temperature trend after 1930 in Fig. 63.3 is far more reliable than the trend before 1925, a hypothesis that is confirmed by a study of the two datasets in question.
Fig. 63.4: The number of station records included each month in the mean temperature trend for Peru when the MRT interval is 1951-1980.
The two stations with data before 1925 have data that is
discontinuous and that fluctuates enormously. One of the two stations is
Arequipa Airport (Berkeley Earth ID:157461) shown in Fig. 63.1 above. The other is Lima-Callao Airport (Berkeley Earth ID:157469).
For the former the temperatures before 1920 are comparable to those
between 1970 and 2000. For the latter they are comparable with
temperatures in the 1960-1980 period. Yet the result in both cases when
this data is combined with the averaged data for 1929 onwards, is to
produce a mean trend for 1900-1920 that is over 1°C lower than the
temperatures seen in the rest of the trend between 1960 and 2000 (see Fig.
63.3). This indicates that the data for these two stations is clearly
inconsistent with the overall trend for the region (compare the data from 1940-2000 in Fig. 63.1
with that in Fig. 63.3), and so the data before 1925 is highly unreliable.
Fig. 63.5: The temperature trend for Peru since 1900. The best fit is applied to the monthly mean data from 1931-2010 and has a positive gradient of +0.16 ± 0.11 °C per century. The monthly temperature changes are defined relative to the 1951-1980 monthly averages.
Now if we compare these result with the results published by Berkeley Earth we once again see a number of major differences. The mean temperature trend becomes less variable and more linear as illustrated in Fig. 63.6 below. The trend in Fig. 63.6 was generated by performing a simple average on the Berkeley Earth adjusted data from the same 42 stations used to generate the temperature trend in Fig. 63.3.
Fig. 63.6: Temperature trend in Peru since 1900 derived by aggregating and averaging the Berkeley Earth adjusted data for all medium stations. The best fit linear trend line (in red) is for the period 1901-2012 and has a gradient of +0.84 ± 0.03 °C/century.
What is clear is that the trends in Fig. 63.6 above are very close to the trends published by Berkeley Earth and shown in Fig. 63.7 below. This comparison clearly shows that a simple average of the adjusted data from the Berkeley Earth data files (Fig. 63.6) gives almost the same result for the regional trend in Peru as the Berkeley Earth version does (Fig. 62.7), even though Berkeley Earth appears to use weighted averages for its regional averaging. This in turn also suggests that weighted averaging is probably not necessary in Peru, and simple averaging of stations is sufficient to generate a reliable trend even though the spread of stations across the county is far from ideal as Fig. 63.2 illustrates.
Fig. 63.7: The temperature trend for Peru since 1860 according to Berkeley Earth.
Clearly there are some significant differences between the temperature trend for Peru based on the original raw temperature data in Fig. 63.3 and that due to the adjusted data used by Berkeley Earth in Fig. 63.5. The exact magnitude of those differences are shown in Fig. 63.8 below.
The effect of the Berkeley Earth adjustments is to reduce the warming after 1990 and to flatten the curve between 1930 and 1950. The rationale for these adjustments is probably to correct for perceived bad data. However, the station frequency plot in Fig. 63.4 suggests that both these adjustments are being applied to data in Fig. 63.3 that should be highly robust, given that it is derived from averaging a large number (over fifteen) of independent datasets. As I have shown previously, averages of more than fifteen stations from the same local region will tend to cancel the errors from each dataset, and so produce a robust and accurate regional trend.
Fig. 63.8: The contribution of Berkeley Earth (BE) adjustments to the BE anomaly data shown in Fig. 63.6 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 1931-2010 has a positive gradient of +0.55 ± 0.06 °C per century. The orange curve shows the contribution just from breakpoint adjustments.
Conclusions
The data in Fig. 63.3 indicates that there has been a sustained but gentle warming of the climate in Peru of about 0.6°C since 1960. As this is the result of averaging between twenty and thirty different temperature records, this warming would appear to be a real effect and not one based on spurious data.
However, the evidence of Fig. 63.3 also stronly suggests that this warming is no greater than the cooling seen before 1960. So overall, temperatures today are no warmer than those of 100 years ago.
The lack of good data before 1930 makes it difficult to assess the significance of the current temperature rise. It could be due to global warming, or it could be due to natural variations.
No comments:
Post a Comment