The Environment Agency (EA) warns of flooding at Frilsham, Stanford Dingley and Bradfield when the level at the Frilsham gauge goes above 62 cm. The EA has no gauge at Stanford Dingley; the Pang Valley Flood Forum (PVFF) runs its own. This page asks a simple question: of the EA gauges on the Pang, which is the best to use to trigger a Flood Warning for Stanford Dingley?

Short answer: the gauge below Bucklebury village (the EA's "Bucklebury gauge", station 2150TH, which the PVFF calls the Bucklebury "Below Village" gauge). It is the only gauge that measures all the water that reaches Stanford Dingley. The Frilsham gauge is upstream of the streams that carry rain runoff from the clay ridge to the south, from Cold Ash to Bradfield, and so is not impacted by large rain events and the EA's new gauge within Bucklebury village is bypassed by the flood channel when the river is high (it's for that reason that this new gauge was implemented specifically to trigger Flood Warnings in Bucklebury itself).

Recommendation: we recommend that the EA's existing flood warning area 061FWF21FrSDBrad (River Pang for Frilsham, Stanford Dingley and Bradfield) should be divided into two (it is already made up of two physical parts). The Frilsham part should be triggered from the Frilsham gauge, and the Stanford Dingley and Bradfield part from the existing Bucklebury gauge.

The level of the river Pang is driven by chalk base flow, combined with runoff following rainfall from the clay ridge running from Cold Ash to Bradfield along the south of the valley. That runoff joins the river as stream, from downstream of Frilsham. There are no streams flowing into the river upstream of Frilsham.

The Environment Agency operates gauges at:

• Frilsham
• at a new level gauge within Bucklebury village on the main channel (which the PVFF calls the "Village (u/s)" gauge, station 2149aTH)
• at the gauging station below Bucklebury village, downstream of where the Bucklebury Flood Alleviation channel rejoins the Pang (the PVFF's "Below Village", station 2150TH)
• and further downstream at Tidmarsh and Pangbourne.

The PVFF operates its own automated level gauges:

• in Bucklebury village ("Village (d/s)")
• at the mill at Stanford Dingley.

There are two relevant flood warning areas: 061FWF21FrSDBrad (River Pang for Frilsham, Stanford Dingley and Bradfield) and 061FWF21Buklebry (River Pang at Bucklebury). The Bucklebury area sits between two parts of 061FWF21FrSDBrad and splits it in two. Warnings for 061FWF21FrSDBrad are triggered at the moment on the basis of the level at the Frilsham gauge.

The only EA gauge before Stanford Dingley that includes the main runoff flow from the clay ridge, is the Bucklebury gauge (station 2150TH).

Using data from the EA's Bucklebury and Frilsham gauges and the PVFF's Stanford Dingley gauge, we have plotted in the charts below: Bucklebury (Below Village) v Stanford Dingley and Frilsham v Stanford Dingley.

Each dot on the two charts below is a Stanford Dingley reading plotted against the level at the other gauge a few hours earlier. Blue dots are the automated Stanford Dingley gauge (from 2024); red dots are the PVFF's manual gauge-board readings from 2014 (those that pass the same Bucklebury filter, below). The level in Stanford Dingley is impacted by whether the mill sluice boards are in or out (and there is no formal record of their position) - so we have filtered only on occasions when the Bucklebury Below Village gauge is above 55cm for both charts (as the boards would not be in such a circumstance). This is an indicator, not a record of the position of the boards. For each we show also correlation r which runs from 0 (no relationship) to 1 (a perfect match); R² is r squared.

Note that initial garden flooding in Stanford Dingley can begin when the level at the mill is 90cm, and real overflow at about 120cm.

The river first broke its bank at the village green in Stanford Dingley at about 11.30 pm on 14 February 2014. The PVFF's manual gauge-board readings had the level at the mill at an estimated 0.95 m at 9 pm and 1.25 m at midnight, so about 1.2 m when the bank broke. The Bucklebury gauge, two hours earlier, read 0.78 m.

• Frilsham has been above the 62 cm trigger on only two occasions since its record began in 2002 (20 July 2007, and 14 and 15 February 2014), and its highest level since the automated Stanford Dingley gauge began (January 2024) is 57 cm, so the current trigger has hardly ever been reached.
• It does not set a trigger level. The automated Stanford Dingley gauge has recorded only one flood reaching 85 cm and none reaching 90 cm; the higher levels are from the manual readings in 2014 (the red dots), which were read by eye from a gauge board. The top of the relationship, where flooding begins, therefore rests on a small number of readings from one winter.
• The two sets of readings come from different conditions. The overflow was lowered in January 2024, after the 2014 readings, so the 2014 relationship may not match today's, and the correlations that include the red dots mix the two. The 2014 readings were also taken at irregular times, mostly around the peaks.
• Works on the floodplain or channels would change the relationship, so any trigger level would need checking after them.

The best way for the EA to know the level in Stanford Dingley, and to have the best gauge to trigger Stanford Dingley flood warnings, would be to have a level gauge in Stanford Dingley itself. Until then, the Bucklebury gauge is the only existing EA gauge that measures the flow reaching Stanford Dingley.

EA gauge levels (15 minute values), the PVFF's Stanford Dingley gauge (hourly, from 27 January 2024) and readings of the Stanford Dingley gauge board (2013 to 2024), all held by the PVFF. The data behind each chart can be downloaded with its download icon.