DARWIN project

DARWIN is the research programme most of the climate work in this garden belongs to. It combines a field observation network across the Galapagos with high-resolution regional climate modelling in WRF, and the two halves exist because neither answers the question on its own.

The question is how water reaches the Galapagos highlands. Sea level is arid, the highlands support cloud forest, and the difference is not explained by rainfall — which is why occult precipitation runs through everything here. Stations tell you what happens where a station is. A model tells you what happens everywhere, if you believe it. The programme is structured around making the second conditional on the first.

What I was trying to establish

• How hydroclimate varies with elevation and season, particularly the garúa season when conventional rainfall is minimal but moisture input is not.
• Whether non-rainfall moisture pathways are large enough to matter ecologically rather than merely being detectable.
• Whether a regional model at island scale reproduces the processes producing those gradients, rather than reproducing the totals by other means.

That last distinction is the one that shaped the modelling. A model can get seasonal rainfall right through compensating errors, and for this question that would be useless — the whole point is the partitioning between rain and fog, so a correct total with wrong partitioning answers nothing.

The field component

The network spans Santa Cruz, San Cristóbal, and Isabela, covering coastal, transition, and highland environments. In broad elevation bands:

• Coastal lowlands, roughly 0-50 m
• The mid-elevation garúa zone, roughly 300-700 m
• Highland crest and crater environments above roughly 700 m

Cerro Crocker is the highland reference site, and the station notes cover what the deployments measure. Public material stays at this level of spatial precision deliberately — the bands are what matter scientifically, and exact siting is not mine to publish.

The modelling component

ERA5 forcing, nested WRF domains configured for tropical island conditions (the tropical setup), with physics choices treated as variables to be tested rather than settings to be inherited — see the experiment suite. The result is the refined analysis, evaluated against the validation dataset.

What came out

The publication thread tracks outputs. The main synthesis is the archipelago-scale refined analysis; heavy rainfall and ecoclimatic cloud zonation are the more specific results, and the ENSO response is the variability that reorganises all of it.

See also: the DARWIN workflow map and the Galapagos climate system map.