With summertime on the horizon, people across Wales are undoubtedly looking forward to saying goodbye to the rain and welcoming warmer temperatures and sunnier skies. However, everyone knows that Wales has volatile and unpredictable weather, even during the summer season, but which Welsh districts have the least dependable summer weather?
Eager to find out, energy switching site SaveOnEnergy collected official data relating to various weather conditions, including rainfall, humidity, wind speed, temperature, and sunshine. We then analysed the data to find out which Welsh districts suffer from the most unpredictable weather and, from this, awarded each district an ‘unpredictability score’ out of 10.
When it comes to all the different weather elements combined, South Wales experiences some of the most unpredictable weather in all of Wales. Of all Welsh districts, nine out of the top 10 districts are situated in the south of the country.
We can reveal Torfaen is the most unpredictable for summer weather, with the district being awarded particularly high unpredictability scores for rainfall, wind speed, and sunshine. The district has been awarded a total unpredictability score of eight out of 10 in our study - the highest of all Welsh districts.
Following closely behind with the second most unpredictable weather during the summer period is Blaenau Gwent. The district scored particularly highly for rainfall and wind speed, resulting in an average total unpredictability score of 7.72 out of 10.
Sir Fynwy, also known as Monmouthshire, has the third most unpredictable summer weather overall, with an average unpredictability score of 7.39 out of 10. The district scores highly for unpredictable temperatures and rainfall during the summer months.
Well known Welsh cities Newport, Cardiff, and Swansea also feature among the 10 districts in Wales, receiving average overall unpredictability scores of 7.06, 6.90 and 6.59 out of 10, respectively.
On the topic of rainfall alone, our study revealed that Neath Port Talbot has the most unpredictable downpours during June to August. In fact, we awarded the district full marks for rainfall (10 out of 10) due to rainfall in the area being particularly volatile.
Rhondda Cynon Taf suffers from the second most unpredictable rainfall throughout the summertime period, with a rainfall unpredictability score of 9.97 out of 10. Closely behind is Pen-y-Bont ar Ogwr scoring 9.94 out of 10 - the third highest score of all Welsh districts studied.
When it came down to relative humidity (how much water is in the air compared to how much it could hold at that temperature), Powys suffers the most from unpredictable humidity, resulting in a score of 9.49 out of 10. The Isle of Anglesey trailed closely behind Powys, with a slightly lower relative humidity unpredictability score of 9.41 out of 10.
The district with the third most unpredictable humidity between June and August is Merthyr Tydfil, with a relative humidity unpredictability score of 9.29 out of 10. Other Welsh districts that feature in the top five include Gwynedd and Blaenau Gwent, with scores of 9.12 and 9.07 out of 10, respectively.
From our analysis of monthly climate observations, we can reveal that Blaenau Gwent is the Welsh district with the most unpredictable wind speed in summer. As a result, the district has been awarded an unpredictability score of 9.8 out of 10. Featuring once again, Powys not only had the most unpredictable humidity during summer, but it also had one of the most unpredictable records when it comes to wind speed during summer too, resulting in a score of 9.66 out of 10. Torfaen has the third most unpredictable wind speed during summer, with an unpredictability score of 9.24 out of 10. This is unsurprising when considering that it is situated close to first place Blaenau Gwent.
From our analysis of monthly climate observations, we can reveal that Blaenau Gwent is the Welsh district with the most unpredictable wind speed in summer. As a result, the district has been awarded an unpredictability score of 9.8 out of 10.
Featuring once again, Powys not only had the most unpredictable humidity during summer, but it also had one of the most unpredictable records when it comes to wind speed during summer too, resulting in a score of 9.66 out of 10.
Torfaen has the third most unpredictable wind speed during summer, with an unpredictability score of 9.24 out of 10. This is unsurprising when considering that it is situated close to first place Blaenau Gwent.
With sunshine being among the most exciting parts of summer, those in the Isle of Anglesey may be disappointed, since the district has the most unpredictable sunshine in Wales throughout the summer season. As a result, the district has been awarded the maximum sunshine unpredictability score, 10 out of 10.
The capital of Wales, Cardiff, has the second most unpredictable sunshine levels between June and August, resulting in a high unpredictability score of 9.88. Bro Morgannwg, near Cardiff, followed in third with an unpredictability score of 9.85 out of 10.
It isn’t only the sunshine that we look forward to when we think about summer, it’s the hot temperatures too. Even though Wales typically have far more reliable and predictable temperatures than England, some Welsh districts still experience more unpredictable temperatures than others.
Sir Fynwy, or Monmouthshire, has the most unpredictable temperatures in Wales during summer, with an unpredictability score of 2.98 out of 10. Other unlucky districts include Casnewydd, with an unpredictability score of 2.95 out of 10. Blaenau Gwent, Torfaen, and Sir y Fflint also featured among the top five, with scores of 2.89, 2.72 and 2.42, respectively.
1. To begin the study, climate data from weather stations interpolated over the UK at a 5km resolution was collected from the Centre for Environmental Data Analysis. SaveOnEnergy collected monthly climate observations (2001 - 2016) alongside monthly long-term averages (1971 - 2000) from the UKCP09 Met Office datasets.
2. Climate variables collected include:
3. To only analyse weather for the summer months, the climate observation data was filtered to only include June, July, and August.
4. In order to reverse geocode the coordinates for the variables collected in step 1 we used the Boundary-Line dataset from the Ordnance Survey. District level boundaries were used as they offered the most granularity. A spatial join was performed to map the coordinates to the district boundaries.
5. The datasets were then aggregated by district and merged.
6. The difference between the long-term average and the observed values were used to calculate the root mean squared error (RMSE) for each variable. The RMSE helps us measure the volatility/unpredictability of the weather variables against their historical averages. Note that comparison to historical averages is more a reflection of volatility than it is of unpredictability. However, it could be argued that volatility is a potential cause of unpredictability. Therefore, this method is justified. A better approach may have involved comparing historical forecasts to observed values, but we were not able to source historical forecasts.
7. Each variable's RMSE was then ranked and totalled to give us a total unpredictability/volatility score out of 10 for each district. This was achieved by adding the score out of 10 for each together dividing it by five (to account for the five variables) to find an average score.