Mountain Rescue Team works for free for our leisure and safety

At the beginning of 2026, an outpouring of donation has been made to Wasdale Mountain Rescue Team to help cover the cost of hotel fee after news of two walkers not paying the hotel after the team rescued them and arranged for them to spend the night in the hotel. However, the cost of the rescue is not limited to the hotel fee. According to the incident report, the rescue involved 8 team members who have spent their evening until after midnight (7.1 hours) to rescue the men. Of course the team did not get paid for their labour, which would be 8 x 7.1 = 57 hours in total, because mountain rescue teams in the UK are operated by volunteers. Besides this recent incident, back in February 2021, a headline reported a mountain rescuer fell for 150m and suffered spinal fractures while out in a mission around midnight to assist a camper with chest pain. Despite these human cost and injuries, the mountain rescue teams continue doing their good work to save life, not to mention the ever-increasing incidents.

To have an overview of the tough work the team has undertaken, I have designed some visualisations to illustrate the team effort in the past ten years (2015-2025).

When do incidents happen?

In the bubble chart below, each bubble represents each incident happening on a specific date in a month (x-axis) and in a year (y-axis). Also shown at the margin of the chart is the sum of the number of the incidents for each year and each month summing across the years. The size of the bubbles is proportionate to the total hours spent in the rescues, that is calculated by number of rescuers times hours spent.

You could hover over them to see more details of each incident and click to be brought to the incident reports and pictures, which would give you more understanding of each incident. For example, the biggest green bubble in May 2024 is an overdue incident that sadly ended in death despite an overnight search involving multi agency team effort.

The bubbles and bars are coloured based on the incident causes

Overlapping and clustering bubbles indicate incidents happened in the same day or within days. This is quite common in summer months (Jun - Aug), however there are also increasing incidents in the winter month (Dec - Feb) since 2021. It should be noted the lack of bubbles at the beginning of 2020 and 2021 corresponds to Covid-19 lockdown that restricted travelling or mingling with others (see also the heatmap below for the total numbers in each month in each year). Although the bubble chart does not allow for precise visualisation of the day that an incident happened, the following bar chart shows that weekends also tend to be busier for the team. This time-wise analysis shows that holidays become the working days for the rescue team.

Injured & Medical and Lost are the common causes (orange and red bubbles) that mountain rescue team was called. However, there are also a few big green bubbles, indicating incidents caused by Overdue have involved more rescuers and hours (10 out of 170 required more than 200 hours). It is likely that Overdue takes more effort and time because the walkers were missing and losing contact with the others, so time is spent locating them. The other three causes are typically reported by the walkers themselves or their company, so it is easier for the team to locate them, although sometimes phone reception and declining battery could cause challenges.

Nonetheless, there are also a few incidents of Injured & Medical and Cragfast that involved large amount of rescue hours, likely due to the complication of reaching the walkers in the Cragfast cases, and the procedures to fetch the injured or sick walkers who could not move themselves from the remote mountain areas. Some of these cases require helicopters. We will visit the association between cause of incident later and time spent and other variables later.

Although Injured & Medical are the most common cause of incidents, there are some months (Oct - Feb ) when there are more Lost incidents. It is likely due to the shorter day time and less visibilities during those months such that people are more susceptible to losing their route.

How long do rescue operations take?

In the gannt chart below, each bar represents an incident in the context of the day it happened to illustrate how long a rescue operation could take as well as when incidents happen in the day. Each grid corresponds to 6 hours. Although UK day time changes across the seasons, it could be safely assumed that 0-6 and 18-24, that is the lowest and highest part of the chart cover mostly the night time. Sometimes, an operation could go on until the next day, that is crossing the next day reference line in the chart.

This is an interactive chart that allows you to pick the year(s) or incident cause(s) to explore the pattern as well as number of operations, average number of operation hours and rescuers involved. For example, by selecting Overdue, and all the years, then you could see that there are 170 operations, with average hours of operation of 3.4 hours and involved on average 6.3 rescuers. Similarly, by selecting all incident causes and year 2020, you could see there were 77 operations with an average of 4 hours of rescue time, and 9.6 of rescuers. You could also select all, and you would see there are 1216 rescues between 2015 and 2025, 20% of them were overnight operations, and on average took 4.1 hours and involved 10 rescuers. The default chart shows all the incidents requiring callout in 2025.

There are two types of callout - Full Callout that involves on average 15 rescuers and takes 4.8 hours; Limited Callout 7 rescuers 4.7 hours. Beside callout, sometimes Alert is logged, when the team is not mobilised, either because it could be resolved on phone or the team leader could not get sufficient evidence or triage it to another mountain rescue team, or the incident is resolved before the team is mobilized. Alert on average involves 1-2 members and takes 2.2 hours, but it could be longer as the team leader would wait or keep track with the walkers until they reached safety. You could see more details about each type of callout and alert by selecting them in the chart.

Overall, although 86% of the incidents occurred after 12 noon, 20% of them going on until the next day and five percent of all the cases started in the unearthly hour before 6am. These overnight and unearthly operations typically require rescuers to walk in the dark, and perhaps cold night, further increasing the difficulty of the rescue.

By selecting both full and limited callout, and investigating each incident cause in each year, it could be observed that Overdue and Lost seems to happen more often in the early morning and late night and rescue during these hours took longer, likely due to the lack of light. In contrast, cragfast seldom happened in early morning and Injury & Medical also mainly happened between 12 noon and 6pm. This has led me to wonder if during day time, people are more daring to explore to the extent that it might lead to be stuck in some unbeaten path and get injured. Furthermore, Cragfast seems to be on the rise since 2023, probably due to more and more amateur walkers attempting difficult areas.

In each year, there seem to be a couple of operations that stretch over two grids, meaning taking 12 hours to complete. This is obvious in year 2015 and 2023. The average number of rescuers involved in callout fluctuate between 12 and 15. These observations really show that incidents could happen at any time in a day and sometimes the rescuers have to sacrifice their sleep to assist the walkers.

Appendix - Statistical Modelling

Seasonality and Trend of Incidents

As observed above, the number of incidents is affected by seasons within a year, and there is also a long term upward trend across the years. To clearly show these patterns in the data, I also ran a statistical modelling -- Seasonal and Trend decomposition using Loess to show these two patterns, as well as to tease out any unexpected number of incidents on top of these patterns. These patterns are shown in the four charts below with the x-axis being month-year:

1. The line in the first chart visualises the total number of incidents in each month across the years, with additional stacked bar to show the breakdown of different incident causes.

2. The line in the trend panel visualises the number of incidents that are associated with the long term trend. It is not possible to show the breakdown of incident cause in this trend, as well as the season and residual below, due to low number of some causes for meaningful statistical analysis.

3. The line in the season panel visualises how the number of incidents change across the season, with peaks usually around July and August, which is the summer months good for hiking.

4. The line in the residual visualises the number of incidents which is not expected by the trend and seasonality, that is there might be some other factors that cause the increase or decrease of the incidents. The two peaks in October in 2017 and 2021 are the odd times that there are more incidents than expected due to seasonality and trend. It awaits further investigation so the team could better prepare in the future for this kind of unexpected rise of incidents.

You can also hover over the line to see the exact number of incidents in each model. It is worth bearing in mind that it is a modelling of the incident number and this modelling essentially breakdown the original number into three components - trend, season, residuals. The number of these three components add up to the original incident number. The negative number in season and residuals is just an indication that there should be fewer incidents due to the season and residuals in those months.

Predicting the future

As observed earlier, the number of rescues has shown steadily increasing trend since 2021 and typically peak in the summer season, that there is seasonality in the number of incidents. Therefore, it is possible to predict the future number of incidents based on the trend and seasonality in the past data. The chart below shows such prediction for the year of 2026, with margin of errors. Prediction would help the mountain rescue team to prepare resources accordingly, but of course let's hope the future number of incidents will go down instead.