What does BAASLAB build in climate AI?

Four systems, all running on our own sites first. Green cover mapping from satellite, to see if planting is growing or dying. Plant performance monitoring, to find losses that never set off an alarm. Land and water stress models, to see where conditions are getting worse. River health monitoring, to locate a change to a stretch and a day.

How is green cover measured from satellite?

Healthy plants reflect infrared light strongly and absorb red light. Comparing those two bands over the same boundary across dates shows whether cover is rising or falling.

Why do you sell measurement at all?

We built it because our own claims are worth nothing without it. Once built, it works just as well on somebody else’s land or plant.

Do you publish your data?

Yes. Anything built from public satellite and weather data is published back with the method and the limits stated.

What we build

Climate AI and IoT

A claim is worth exactly what its measurement is worth. So we build the measurement, publish the method, and treat an unverifiable number as no number at all.

Still water with woodland along the far bank

Why this exists

Measurement is the product

Credits have been sold against plantations that did not survive
Buyers now discount everything until it is proven
A tonne you can verify is worth more than one you assert
That price gap is wide enough to build a business in

So it runs on every project we do, and it is included rather than sold as an extra.

Four systems

What each one answers

Green cover

Is the planting growing, holding, or quietly dying between site visits?

Plant performance

Is this plant producing what the weather should have given it, and if not, where is it going?

Land and water stress

Which districts are deteriorating fastest, and how much warning is there?

River health

Which stretch of the river changed, and on which day?

See how

In detail

River health monitoring

One station tells you a river is dirty. Everyone already knows that.
A pair, above and below a discharge, tells you which stretch and which day
That turns a complaint into evidence
Cleaning is measured in years, so continuous beats sampling

Continuous river monitoring, in outline. The design decision that matters is pairing. One station tells you a river is dirty, which everybody already knows. A station above a discharge and another below it tells you which stretch changed and when, and that turns a complaint into evidence. Cleaning a river is measured in years, and without continuous readings there is no honest way to say whether the money spent on it worked.

Six services

Measurement, as work you can buy

MRV design

MRV protocol document with methods, frequencies and verification route

Scope

Remote sensing and green cover

Green cover time series with change detection over the defined boundary

Scope

Plant performance monitoring

Performance ratio reporting with loss attribution and alerting on divergence

Scope

River health monitoring

Continuous river health index with divergence alerts and an evidence record

Scope

Soil and carbon accounting

Soil carbon baseline and change reporting with method stated

Scope

Reporting and independent verification

Verified reports and public disclosure with the method attached

Scope

What measurement changes

Price Proven tonnes sell higher Buyers pay more for removal they can defend publicly, and discount everything they cannot.

Operations Losses found early Soiling, dead strings and tracker drift cost generation daily and set off no alarm.

Public money Programmes can be judged Continuous river data shows whether money spent on a stretch actually changed it.

Trust Claims become checkable Published methods and published bad results are what separate a record from a brochure.

In depth

Why measurement decides the price

Mechanisms

All articles

Questions