Lisaris EO Studio
Axine Labs' Earth-observation analytics platform. Draw an area; it measures what
is on the ground there from satellite imagery, and states how far each number
can be trusted.
What it measures
- Mining & minerals: alteration mineralogy, ore-associated
minerals, acid drainage, tailings.
- Energy: hydrocarbons on soil and water, and methane plumes.
- Agriculture: crop vigour, canopy chlorophyll and leaf area,
evapotranspiration, moisture, residue.
- Insurance: event response and exposure, including burn
severity, flood extent and built-up change.
- Forestry: canopy condition, narrowband pigment, stress.
- Water: water extent, moisture, sediment, algal pigment.
- Land cover & built environment: what is on the ground,
and what has been built on it.
Imagery
Scene search runs over the public archives: Sentinel-2, Landsat 8 and 9,
EMIT, Sentinel-3 OLCI and PACE OCI. Searching and browsing are free and need
no account. Where a public sensor cannot answer the question, a new tasked
capture can be quoted.
Plans
A plan decides which analytics you can run and over how much ground, with
no per-run charge. The free Standard plan includes the band-arithmetic layers;
Professional and Enterprise include everything. Orders are separate from plans
and buy imagery only.
How far to trust a number
Every layer carries a validation status. Deterministic is exact,
repeatable arithmetic. Screening is an established method applied
without local calibration: compare patterns within a scene, and treat absolute
values as indicative. Regional carries a measured accuracy inside
calibrated regions only. Not scored is unsupervised, with precision and
recall undefined.
No accuracy, detection rate, detection limit or emission rate is
claimed for anything this platform computes. Where such a figure would
normally appear you will find the validation status instead.
This page
Below is the HTTP API reference, for teams integrating Lisaris into their own
systems. It needs JavaScript to render. Most people use Lisaris through the web
application instead.