Florida's 2026 freeze is visible from orbit in citrus and blueberries, and almost invisible in strawberries

What happened

Winter Storm Ezra (30 December 2025 to 1 January 2026) and Winter Storm Gianna (26 January to 4 February 2026) brought freezing temperatures to 66 of Florida's 67 counties. The state agriculture department's preliminary estimate (20 February 2026) puts losses above US$3.1 billion. It reports roughly 80% of remaining strawberry production and 90% of remaining blueberry production lost, and about 80% of citrus acreage significantly affected. USDA issued a disaster declaration on 4 March 2026.

What we did

We mapped strawberry, blueberry and citrus fields from the national crop map in three bounded areas: the Plant City–Dover strawberry belt (250 km²), the blueberry cluster around Windsor in Alachua County (200 km²), and the citrus belt around Avon Park, between Lake Wales and Lake Placid (300 km²).

For every field we built a weekly vegetation record at 20 m from October 2022 to May 2026. We compared the weeks after the freeze with the same weeks of the three previous seasons. We counted the hours each field spent below its crop's damage threshold, using the research weather-station network corrected night by night with satellite land-surface temperature.

A field counts as damaged only if it lost green canopy beyond what ordinary year-to-year variation produces, and was exposed to damaging cold. The cut-offs were set so that, in the three previous seasons, no more than about 5% of fields would have been flagged. They were not tuned to the state's numbers. Results are rolled up by area and by county.

What we found

Citrus: 76% of mapped grove area shows freeze damage (32% severe) in the Avon Park area. The state's figure is about 80% of citrus acreage significantly affected. Damage is heavier in the Polk County part of the area (90%) than in the Highlands part (71%). Polk was colder: a median of 9 hours below −2 °C at 2 m, against 5 hours.

Blueberries: 44% of mapped blueberry area is damaged (38% severe) around Windsor. Every block spent more than a day below −2.2 °C at 60 cm (median 31 hours). That is half the state's 90% production loss, but close to the 40–45% one Alachua grower reported for their own farm. Losing the crop does not always mean losing the canopy. We found no post-freeze hedging signal in February to April; the same blocks showed clear summer pruning in June 2025.

Strawberries: every field was exposed, almost none lost canopy. All 219 mapped fields saw at least 6 hours below −2 °C at 60 cm (median 14 hours), yet their vegetation index ran slightly above normal after the freeze. Strawberry freeze loss is flowers and fruit, under ice from freeze-protection irrigation; canopy greenness cannot see it. A yield model built on canopy greenness and frost hours has no skill for Florida strawberries. Strawberry losses need a different signal: harvest records, bloom counts or very-high-resolution imagery.

Limits

Acreage affected is not production loss. The state reports production losses for berries and an acreage share for citrus, statewide; ours are crop-area shares inside three bounded areas.

20 m vegetation data is field-scale, not row-scale; field edges mix with roads and bare soil. Winter cloud left only 62–75% of post-freeze field-weeks usable. Citrus damage expresses over weeks and our window ends 29 March, so late damage is missed and spring flush can mask it.

Field boundaries are derived from the national crop map, not surveyed: 10–52% of mapped crop area falls below the 1 ha minimum or is lost in shape clean-up, and large grove complexes are split into ~25 ha units. The crop map under-maps Florida blueberries, so the blueberry result rests on 12 fields (46 ha). Frost hours are interpolated from 21 stations, with a typical error of 8–14 hours per site; cold pockets smaller than 1 km are not resolved.

Validation