August was cooler than last August across the Sacramento Valley and pulled more water out of the trees anyway. For most of the month, the plant signal and the weather signal pointed in opposite directions.
We read 318 California blocks through August — 9,385 block-days across eight crops. Our Sacramento Valley almond blocks bottomed out in the second week at −8.9 Δbar, then recovered three full bars by the end of the month. Reference ET across the middle of that recovery did not fall. It rose slightly.
That gap is the month in one sentence, and it is the reason we publish this index.
What a Δbar is
Every number below except the apple and citrus lines is a Δbar: the distance between what a tree measured at midday and what a fully watered tree would have measured under that block’s own weather, that same day. Zero means the tree kept up with the day. −8 means it finished eight bars behind. We report it this way because a raw stem water potential reading moves when the weather moves, and we want a number that compares a block to itself in June and in September.
Bands are our published scale: no stress above −2, minimal to mild from −2 to −6, moderate from −6 to −8, high stress or beyond at −8 and drier.
The almond network
Every figure below aggregates block averages, not individual sensors. A block is the unit a valve actually controls.
| Region | Blocks | August avg | Median | Driest tenth | Wettest tenth | High stress or beyond |
|---|---|---|---|---|---|---|
| Sacramento Valley | 133 | −7.7 | −7.2 | −15.0 | −1.4 | 43% |
| Northern San Joaquin | 67 | −6.3 | −5.8 | −11.0 | −1.2 | 28% |
| Southern San Joaquin | 6 | −4.5 | −3.0 | too few blocks to call a network reading | ||
The Sacramento Valley average hides a valley that is nowhere near itself. The driest tenth of those 133 blocks sat at −15.0 Δbar for the month; the wettest tenth at −1.4. Thirteen and a half bars of spread, same crop, same valley, same weeks.
The week the two signals came apart
| Week of | Sac Valley almond (Δbar) | High stress or beyond | Reference ET (mm/day) |
|---|---|---|---|
| July 27 | −7.9 | 45% | 8.47 |
| August 3 | −8.5 | 57% | 8.55 |
| August 10 | −8.9 | 53% | 7.29 |
| August 17 | −7.4 | 44% | 7.42 |
| August 24 | −6.1 | 33% | 7.53 |
| August 31 | −5.9 | 29% | 6.17 |
Look at the middle of that table. Between the week of August 10 and the week of August 24, reference ET went up — 7.29 to 7.53 mm/day. The trees went the other way, by 2.8 bars, and the share of blocks in high stress or beyond fell from 53% to 33%.
A schedule driven by evaporative demand would have read those two weeks as near-identical. The orchards were not. Something other than the weather changed, and it changed a lot.
The obvious candidate is water going back on after shaking. We will not state that as fact, because our stage tags do not support it yet: 118 of the 133 Sacramento Valley blocks still carry a Hull Split tag, the most recent set on August 12, and only three read Post-Harvest. Those tags are grower-maintained and they lag. What the sensors establish is the shape and the size of the move. Why is a separate question.
The same blocks, one year apart
Year-over-year numbers are misleading when the network grows, so this table holds only blocks we read in both Augusts.
| Crop and region | Blocks | Aug 2025 | Aug 2026 | High stress or beyond, 2025 → 2026 |
|---|---|---|---|---|
| Almond, Sacramento Valley | 62 | −6.4 | −7.0 | 29% → 42% |
| Almond, Northern San Joaquin | 26 | −4.1 | −5.1 | 15% → 15% |
| Prune, Sacramento Valley | 17 | −5.4 | −5.7 | 29% → 41% |
| Wine grape, North Coast | 20 | −5.8 | −5.4 | 15% → 20% |
The Sacramento Valley almond average moved just six tenths of a bar. The tail moved much further — 29% of those blocks spent August in high stress or beyond last year, 42% this year. When an average barely moves and the tail grows by thirteen points, the average is the wrong number to be watching.
Here is the part we did not expect. The same 78 Sacramento Valley weather stations recorded a cooler August this year and a thirstier one.
| Same 78 stations | Aug 2025 | Aug 2026 |
|---|---|---|
| Reference ET | 7.22 mm/day | 7.93 mm/day |
| Average daily high | 100.2°F | 95.8°F |
| Mean relative humidity | 49.4% | 41.6% |
| Solar radiation | 267 W/m² | 283 W/m² |
| Peak wind | 4.82 m/s | 5.10 m/s |
4.4°F cooler, and 10% more evaporative demand. The physics is straightforward once you look past the thermometer: relative humidity fell 7.8 points, sunlight rose 6%, wind rose 6%. Drier, brighter, breezier air pulls water out of a leaf whether or not the day feels hot.
An operation that judges the season by how hot it felt would have called this August the easier one.
The monthly average hides the month
We measured how far each Sacramento Valley almond block travelled within August — its wettest week against its driest. The median block moved 8.1 bars. 114 of 133 moved at least four. And 70 of them, more than half the region, crossed into high stress and back out again inside the same month.
Every monthly number in this report, then, describes a place those trees passed through rather than a place they sat.
Wine grapes went the other way
Our 56 North Coast wine grape blocks averaged −5.4 Δbar for August, median −5.6, with 30% of blocks in the moderate band and 16% in high stress or beyond. A tighter, more uniform group than almonds — the driest tenth was at −9.2, not −15.0.
The direction is the interesting part. As almonds relaxed through the second half of the month, the vineyards did the opposite: the share of blocks in high stress or beyond ran 11%, 16%, 21%, 27%, 24%, 25% across the six weeks from July 27. The vineyards tightened while the orchards let go.
No surprise once you say it out loud — the two crops sit at opposite ends of their seasons, one coming off harvest and one heading into it. But it does mean a statewide number averaged across crops would have shown a flat, quiet August and missed both halves of it.
Prune, and a split fleet
Eighteen Sacramento Valley prune blocks averaged −5.5 Δbar, median −6.1, and ended the month with a sharp release: −6.1 in the first full week of August down to −3.4 in the last, the high-stress share falling from 44% to 17% as post-harvest tags began appearing.
The distribution is a barbell rather than a bell. A third of these blocks spent August at no stress at all; 39% spent it in high stress or beyond. There is not much of a middle, and on a fleet shaped like that the mean is close to meaningless. The split is the only honest summary.
Apple, olive, citrus
Not enough California coverage to publish a network reading. We read three California apple blocks, two citrus and one olive this month. The apple blocks averaged −16.9 bar absolute midday, which is a stressed number, but three blocks is an indication and not a measurement of anything larger. Our apple network is concentrated in New York, and this is a California report.
What we are watching in September
Demand drops off a cliff in the first ten days. Across all 138 Sacramento Valley stations, reference ET falls to a forecast 4.12 mm/day from an August average of 7.71, with highs near 79°F. The Northern San Joaquin runs 4.72, the Southern San Joaquin 5.83, the North Coast 3.45.
For almond and prune blocks coming off harvest, that falling demand lands on the post-harvest window — the trees still have leaves, and while they do they are building next season’s spurs and buds on this season’s water. We went through the published evidence on that window, including the Goldhamer and Viveros work on post-harvest deprivation, in our post on post-harvest irrigation.
The blocks worth watching are the ones that have not turned. At the end of August, 29% of Sacramento Valley almond blocks were still in high stress or beyond, three weeks past the network’s bottom. Some are simply late. Some are blocks where the water went back on and did not arrive the way the grower thinks it did — which shows up in the plant now rather than in a yield report next August.
On the vineyard side, September is harvest, and the North Coast was still tightening when the month closed.
FloraPulse measures stem water potential continuously from inside the trunk, block by block. See what California orchards and vineyards are reading this week.
About the data: 318 California blocks, 9,385 block-days, August 1–31 2026, from the FloraPulse sensor network — continuous microtensiometer measurements inside the trunk. Block averages only, anonymized and aggregated by region; North Coast covers Napa and Sonoma and is reported separately from the Central Valley. Weather figures are per-station averages, with year-over-year comparisons restricted to stations reporting in both periods.