In-situ datasets platform — WQ × Sentinel-2

Value distribution

Measurements over time

Top feature correlations (|Pearson R| vs in-situ)

S2 match quality

Colour by Size by Log colour scale

Type Colour by Band overlay Log Y Rolling mean

X axis Colour by Trendline + R² Log X Log Y

Mode Log Y Reflectance vs wavelength (443–2190 nm)

Chart Variable Group by Log scale

Method Ranking of every band & index against the in-situ value, plus a feature×feature matrix.

Feature → value ranking

Feature correlation matrix

Chart X axis Y axis (left) 2nd Y (right) Group / colour Aggregate Trendline + R² Log X Log Y

Build any chart from the current filtered selection — compare any two columns, add a 2nd field on the right axis, group by a category, aggregate, add a trendline. Honours all active filters. Download gives a slide-ready PNG with a solid background.

Mode

Format

Rebuild your own CSV from the mother datasets — tick exactly the columns you want, optionally group by a category and aggregate the numbers. Honours all active filters.

Columns

Preview (first 12 rows)

Rows/page

1 · Scope

Current filtered selection (0 rows) Entire dataset (0 rows)

ML presets

Keep only standard QC rows Drop rows with missing bands / value Min-max normalise band & index columns (0–1) Add random train/test split column (80/20)

2 · Columns

3 · Format & download

Format

Preview (first 8 rows)