Since when did hydrologists become historians? Matthew Fry and Hollie Cooper from UKCEH’s hydrological data team share the process of unearthing 35 years of data in mid-Wales and how this enables hydrologists to understand river processes.
What are the Plynlimon Research Catchments?
The catchments were established as an open air laboratory in mid-Wales in the late 1960s by UKCEH across the sources of the Wye and Severn rivers. The research catchments answer pressing questions of hydrology, in particular how upland trees affect water resources, informing a huge number of papers and reports. However, the baseline measurements of meteorology and river flows haven’t been publicised – until now!
What data is being made available and why is it exciting?
The dataset comprises over 35 years of hourly weather data (rainfall, temperature, humidity, radiation, wind speed) from six sites within the Plynlimon catchments, and 15 minute river flows from 10 flow gauging stations, from the mid-1970s to 2010.
The data is exciting for a few reasons:
The level of detail: Having such long-term, high-resolution records for multiple locations across a small (<20km2) area is really rare. They enable detailed insight into river flow generation processes across grassland (Wye) and forested (Severn) catchments.
Data accuracy: Many of the weather station locations ran two separate systems in tandem for up to five years, providing unprecedented data to confirm the validity and uncertainty of these measurements. Rainfall was measured by ground-flush gauges to reduce the impact of wind on measurements . The dataset complements nearly 30 years of weekly chemistry sampling from rainfall and streamflow within the catchments, allowing us to understand the dynamics of flood events in some of the wettest areas in the UK.
Continuing a legacy: Importantly, measurements in the area will be enhanced as the Upper Severn becomes an observatory within the Floods and Droughts Research Infrastructure, continuing and enhancing data availability, allowing us to see changes over time and answer new questions. This is especially important as the effects of climate disruption become more visible.
What was the process of bringing together all that historic data?
Gathering the data required a “data rescue” effort and a lot of detective work – including finding (and quality controlling) the raw data. Fortunately, this archive was based on years of best practice at UKCEH, and the original Oracle database tables are still accessible. Sharing data with other expert users of the catchments was incredibly useful. We’re grateful to Professor James Kirchner of ETH Zurich, who put in a huge amount of time into checking through, correcting and removing data.
How has the way we monitored and analysed data at Plynlimon changed?
The instruments used and the way the data is captured, stored and processed, have completely changed over the lifetime of the Plynlimon catchment research.
Originally, paper charts were used to record water levels by tracking a float in a well connected to the river. Operating this equipment in very tough environments on the top of a mountainside brought a lot of challenges and innovations. Plynlimon was at the cutting edge in the development of automatic weather stations and the move from punch tape to cassette tapes through to modern solid state data loggers.
A team of researchers and field technicians used to live close by and visit the catchments daily to take measurements. However, now, sensors in the FDRI project are transmitting one-minute resolution in real time for field teams to check remotely - an incredible development in cost-efficiency and data resolution. Understanding and documenting the changing technology has been a data rescue activity in itself!
How can the dataset be used and where can you access it?
Alongside existing published Plynlimon data, the datasets have diverse research and applied uses. This includes: high-resolution hydrological modelling, analysis to understand flood and drought events and how different land uses affect flood event dynamics.The dataset is available from the Environmental Information Data Centre via this link. Data from 2010 to present, as well as other Plynlimon datasets including the network of >30 monthly storage rain gauges (1970-2010) and daily manual weather station measurements (1970-2000) will be published over coming years once they have been reviewed and quality controlled.
Acknowledgements
A large team over several decades has made the publication of this dataset possible. The authors would like to thank every team member who has helped install and maintain sites; collect and store data; and process, visualise, and analyse the datasets, including (in alphabetical order): Nuria Bachiller Jareno, Jim Blackie, Emma Cowley, Charles Eeles, Simon Grant, David Norris, Anne Roberts, Gareth Roberts, Oliver Swain, Roger Templeman, Alan Warwick, and many others. The authors would like to acknowledge in particular John Smart and Jim Hudson for managing the complex issues involved in collecting and maintaining data quality in often physically challenging conditions in remote areas.