A wet June measured against five years of my own figures

01

What June 2026 actually did

The total for the month was 118.5 mm against a six year June average on this plot of 52.0 mm. That is more than double, and it is the wettest month of any name in my record apart from October 2023. Twenty one of the thirty days recorded measurable rain, which I define as 0.5 mm or more in the morning reading.

The distribution is where it gets interesting. Two days carried a third of the month: 21.0 mm on 8 June and 17.5 mm on 22 June, both from thunderstorms in the late afternoon that I watched arrive from the shed door. Another eleven days recorded between 0.5 and 2.0 mm, the sort of rain that wets the surface, keeps you off the plot and does almost nothing for a bean root at 200 mm down.

Temperature went with it. The mean of my daily maxima for the month was 18.2 C against an average of 21.0 C in the five previous Junes, and there were only four days above 22 C all month. The soil thermometer at 100 mm did not get above 16.0 C until 26 June, which in a normal year here happens in the third week of May.

• June 2026 total, 118.5 mm; six year June mean, 52.0 mm
• Wettest day, 21.0 mm on 8 June; second, 17.5 mm on 22 June
• Twenty one rain days of 0.5 mm or more out of thirty
• Mean daily maximum 18.2 C, against 21.0 C for the five Junes before

02

Against the five Junes before it

The June totals since I started reading properly run 38.0 mm in 2021, 24.5 mm in 2022, 61.0 mm in 2023, 44.5 mm in 2024 and 92.0 mm in 2025. So 2026 is not a freak on its own; it is the second wet June in a row, and 2025 was already well above anything in the first four years. Two years is not a trend and I am not going to pretend it is one, but it is enough to have changed what I do.

What the five year comparison shows most clearly is that the number of rain days moves much less than the total. Rain days for those five Junes were 14, 9, 17, 13 and 19, against 21 this year. The total has swung by a factor of nearly five while the count of wet days has not quite doubled, which means the variation is almost entirely in how heavy the wet days are.

Those two ways of counting are why the site conversation and the gauge disagreed. People remember a wet month as one where they kept getting rained off, and by that measure 2023 and 2026 feel similar. By total, 2026 delivered nearly twice what 2023 did. Both descriptions are true and they are answering different questions.

03

Rain days are not rainfall

I now keep a separate column for what I call useful rain, meaning days of 5 mm or more, because that is roughly the threshold at which water reaches below the top 50 mm of my soil rather than evaporating off it. In June 2026 there were six such days, which is high but not extraordinary. In the drought June of 2022 there was exactly one.

This distinction has saved me a lot of wasted effort. A fortnight of drizzle can leave the beds visibly wet on top and bone dry at root level, and if I go by appearance I do not water and the runner beans sulk. Twice in 2024 I dug a small hole with a trowel next to a row that looked thoroughly wet and found dust at 80 mm. Now, in any spell of light rain, I check with a trowel rather than with my eyes.

The reverse trap is real too. After 21.0 mm in one afternoon on 8 June the beds were saturated to depth, and I stayed off them entirely for four days, because walking on this loam when it is that wet closes it up and I pay for it for the rest of the season. The gauge tells me when not to work as often as it tells me when to water.

04

What the soil did that the gauge did not show

The bottom two beds held standing water in the wheel ruts for six days after 22 June, which has happened twice before in the record, both times in winter. In June, with everything growing, that did visible damage: the carrots at that end went to a yellow colour and never really recovered, and I lifted a shorter crop in October than any year since 2021.

The soil thermometer explained part of it. A saturated soil at 14 C is a cold, airless place for roots, and the difference between the top bed at 16.5 C and the bottom bed at 13.5 C on 25 June was as much about water content as about the slope. I had always thought of the bottom beds as merely damp. They are actually several degrees colder for weeks at a time in a wet season.

The remedy is drainage work I have been avoiding for four years, and over this winter I have finally cut a shallow channel along the bottom edge to the hedge gap. It is not elegant. But I have six years of figures showing that the bottom of this plot is a different growing environment from the top, and the record is what finally made me act on it.

05

What the record changed about sowing dates

The first change is that I no longer sow to the calendar at all for anything that goes directly into the ground. The trigger is the soil thermometer: 10 C at 100 mm for three consecutive mornings for peas and broad beans, 12 C for beetroot and chard, 14 C for French beans and sweetcorn. In 2026 that pushed the French beans back to 21 June, three weeks later than my old fixed date, and they still cropped from the second week of August.

The second is that I have split the main sowings in two, with about twelve days between them, for beans, beetroot and salad. In four of the last six seasons the two batches have ended up cropping within a week of each other, because the later sowing caught better conditions and grew faster. In 2026 the later batch was clearly ahead, and in the dry year of 2022 the earlier batch won easily. Splitting costs seed and nothing else.

The third change is about brassicas and standing water. The bottom beds no longer carry anything that has to stay in the ground through the winter, because the figures show that end waterlogged for long spells in four of the six years. That bed is now summer salads and early potatoes, which are out of the ground by August. None of these are original ideas. They are simply what six years of my own numbers made undeniable.