WRSI vs VCI: which water-stress signal belongs in your trigger calendar

WRSI and VCI keep showing up in the same breath on food-security working groups, usually right before someone asks which one should set off the trigger. They're not measuring the same thing, and mixing them up in a calendar note is a fast way to argue past each other in a meeting.

What WRSI measures

The Water Requirement Satisfaction Index runs a water balance model against a specific crop's growing cycle. It needs a crop calendar (planting date, length of season, growth stages for that variety), a soil water holding capacity estimate, rainfall estimates, and potential evapotranspiration. The output is a percentage: how much of the crop's water demand got met through the season, weighted so that a deficit during flowering counts more than the same deficit during early vegetative growth.

That crop-specific anomaly framing is WRSI's whole value. A 20% rainfall deficit in week three of a 90-day sorghum cycle doesn't hit the same way as the same deficit in week seven, and WRSI reflects that. The cost is that it's only as good as its inputs. If the rainfall estimate is biased for your basin, or the crop calendar doesn't match what's in the ground that season, the water requirement satisfaction index output inherits those errors quietly.

What VCI measures

The Vegetation Condition Index skips the crop model entirely. It takes NDVI for a pixel, compares this month's value against the historical minimum and maximum for that same pixel and that same time of year, and scales it to a 0-100 range. No crop calendar, no soil data, no evapotranspiration model. It's a straight read of how green things are relative to how green they've been.

That makes VCI cheap to run and hard to break. It doesn't care whether your crop calendar is three weeks off, because it isn't using one. But it also doesn't know if it's looking at a cropping zone, a grazing zone, or scrubland recovering from last year's fire. Vegetation stress from a pest outbreak or a late planting looks the same in a VCI anomaly as vegetation stress from water deficit. It tells you something is off. It doesn't tell you why.

Where the two disagree, and why that's useful

A zone can show a weak WRSI, poor water satisfaction for the specific crop modeled, while VCI stays near normal, often because the broader land cover isn't under the same stress as the modeled crop. The reverse happens too: VCI drops sharply while WRSI holds steady, usually a sign of something other than rainfall deficit driving the vegetation anomaly, like a heat spell outside the water balance model's scope or a pest event moving through.

For a cropping zone where you have a crop calendar you trust, WRSI is the sharper tool because it's built for exactly that crop and that stage. For a grazing zone, or anywhere the crop calendar is shaky or the land cover is mixed, VCI's lack of agronomic assumptions stops being a weakness. It reads the vegetation as it is, not as a model assumes it should behave.

Feeding either one into a trigger decision

Neither index is the trigger on its own. Both are inputs that need a threshold and a comparison to the seasonal norm before they mean anything for an anticipatory action calendar. Pick the one that matches what you know about the zone: a crop calendar you trust points to WRSI, mixed land cover or a shaky crop calendar points to VCI, and tracking both lets you flag the divergence itself as a signal worth a closer look.

Either way, the monthly pull is the same chore: get this month's value, get the seasonal norm for that pixel and month, and get the gap between them onto the calendar before the trigger date. Food Security Warning delivers that comparison as a single anomaly number per zone, pulled monthly from coarse-resolution satellite, so the input sits in the calendar instead of in a side project to rebuild it by hand.

If your calendar still runs on a manual pull-and-plot, it might be worth seeing what a monthly anomaly feed looks like for your zones.