GIS Tools - Raster Value Probe

What Is a Raster Value Probe?

A raster value probe is an interactive map tool that lets you click anywhere on a raster layer and instantly read the underlying pixel value. For a Digital Elevation Model the probe returns elevation in metres; for a Sentinel-2 reflectance band it returns a 16-bit integer; for a land-cover classification it returns a category code; for an NDVI raster it returns a value between -1 and 1. The probe is the GIS equivalent of the eye-dropper tool in an image editor — a fast, low-friction way to interrogate continuous and categorical raster data without scripting.

The Raster Value Probe on gis.tools loads a GeoTIFF in the browser and displays the pixel value, plus the geographic coordinates and the row/column index, in a popup or fixed panel as your cursor moves across the map. It is the simplest possible raster query tool, and yet it is one of the most useful: you cannot trust a raster you have not poked. People search for "raster query tool", "click raster value", "sample raster pixel" and "raster identify gis" because every desktop GIS provides this functionality and users expect it from any browser raster viewer.

Unlike a Spatial Filter or Feature Identify Tool, which return vector attributes, the probe returns the raw raster cell value at the cursor location, applying the inverse of the affine transform to map screen pixels back to raster row/column indices.

How the Probe Works

When you click the map, the tool performs three coordinate transformations and one array lookup.

Step 1: Screen to Map Coordinates

The map library converts the screen pixel under the cursor to map coordinates in the basemap CRS (usually Web Mercator EPSG:3857 or WGS84 EPSG:4326).

Step 2: Map to Raster CRS

If the raster is in a different CRS, the map coordinates are reprojected to the raster's native CRS using a Proj4-style transformation.

Step 3: Raster CRS to Row/Column

The raster's affine transform — six numbers stored in the GeoTIFF header — converts the map coordinates to fractional row and column indices. The standard transform is x_map = a + col * b + row * c; y_map = d + col * e + row * f. The probe inverts this to get row and column.

Step 4: Pixel Lookup

The probe reads the typed array at index (row * width + col) for the requested band and returns the value. For multi-band rasters, it returns one number per band. If the value matches the nodata sentinel, the probe reports "nodata" instead of a number.

Subpixel Sampling

Some probes offer bilinear interpolation between the four neighbouring cells, which is appropriate for continuous data like elevation. Categorical data (land cover, soil class) must always use nearest-neighbour to preserve integer codes.

Key Parameters and Display Options

Sampling Method

Nearest-neighbour returns the value of the cell that contains the cursor — correct for categorical data. Bilinear returns a weighted average of the four nearest cells — appropriate for continuous data and produces smoother readouts when zooming.

Coordinate Display

Choose between map coordinates in the basemap CRS, raster-native coordinates, latitude/longitude in WGS84, UTM, MGRS or any custom Proj4 string. Use the Coordinate Format Converter for batch conversions.

Multi-Band Output

For RGB and multispectral rasters, the probe shows all band values in a list. For NDVI products derived from band math, the probe shows the computed result.

Nodata Handling

The probe explicitly reports "nodata" when the queried cell matches the nodata sentinel (typically -9999, NaN or 0).

Hover vs Click Mode

In hover mode the readout updates continuously as the cursor moves. In click mode the value is recorded only on tap, which is better on mobile or for dense rasters where streaming reads would slow the page.

Practical Applications

Quick Elevation Check

A hiker planning a route opens a DEM tile in the browser and probes the highest point of a ridge to confirm the summit elevation. A pilot opening a sectional chart raster probes ground elevation under a glide path. A surveyor compares a GNSS measurement against a published DEM to assess vertical accuracy.

Sentinel-2 Reflectance Inspection

A remote sensing analyst clicks a suspected water body in a Sentinel-2 NIR band to check whether the reflectance is consistent with water (low NIR) or wet vegetation (moderate NIR). Probing a few representative pixels is the fastest way to validate a classification before running zonal statistics.

Land Cover Validation

After loading a national land cover raster, an ecologist clicks 50 known field-verified points and compares the probe output to ground truth. Mismatches reveal classification errors and inform a confusion matrix.

NDVI Sanity Check

After running the NDVI Calculator, the analyst probes a known healthy forest pixel (expected NDVI 0.7-0.85) and a known asphalt pixel (expected close to zero). If the values match expectations, the calculation is correct.

Soil and Geology Lookup

A field geologist probes a soil class raster to confirm the mapped unit at a sampling point matches the field description. Discrepancies trigger a re-survey.

Hydrology and Flood Risk

An emergency planner probes a flood depth raster at the address of a vulnerable building to estimate worst-case water height during a 100-year storm.

Climate and Environmental Variables

Climate scientists probe gridded temperature, precipitation, evapotranspiration or solar irradiance rasters at city locations to extract point values for downstream modelling.

Step-by-Step Workflow in gis.tools

Open the Raster Value Probe page on gis.tools.
Drag a GeoTIFF (single or multi-band) into the page. The file is decoded locally with geotiff.js.
The raster overlays on the basemap at its georeferenced bounds.
Click or hover anywhere on the raster. A popup shows the value(s), the geographic coordinates, the row/column index and the cell size.
Switch the coordinate display to UTM, MGRS or DD/DMS as needed.
Toggle nearest vs bilinear sampling depending on whether the data is categorical or continuous.
Combine with the Hillshade Generator, Slope/Aspect Calculator or Zonal Statistics for richer interrogation.

Worked Example: Validating a Sentinel-2 Cloud Mask

You have a Sentinel-2 Level-2A scene over Sicily. The scene includes a Scene Classification Layer (SCL) raster with integer codes (0 = no data, 4 = vegetation, 6 = water, 8 = medium cloud, 9 = high cloud). You drop the SCL band into the Raster Value Probe and click on a bright white area in the visual band that you suspect is a cloud. The probe returns 8 — confirmed medium cloud. You then click a dark patch over Mount Etna and get 4 — vegetation, which surprises you. Looking closer you realise the area is shadow from a cumulus cloud miscategorised by the SCL. You note this and decide to mask SCL value 3 (cloud shadow) more aggressively in your downstream NDVI workflow.

Common Pitfalls and Gotchas

Bilinear sampling on categorical data produces nonsense — always use nearest neighbour for class rasters.
Probing a coordinate outside the raster bounds returns nodata, not zero — make sure your basemap zoom matches the raster extent.
Multi-band rasters return one value per band; the probe popup may be wide. Zoom in for accurate clicks.
If the raster CRS is not properly tagged, the probe will report wrong coordinates. Check with the CRS Metadata Inspector.
Floating-point rasters with NaN nodata may display "NaN" rather than "nodata" — both mean the same thing.
Edge pixels (especially after reprojection) often contain nodata or fill values; expect a transition zone.
Coordinate display precision matters — for sub-metre rasters, use 6 decimal degrees or full UTM precision.
Mobile touch events register over a wider area than a mouse click; use the click-locked mode for precise sampling on touch devices.

Tips for Best Results

Cross-validate probe values against an independent source (GNSS, ground truth, published statistics) before trusting a raster.
For continuous rasters use bilinear sampling; for categorical rasters use nearest-neighbour.
When probing many points in sequence, copy the readouts into a CSV via the Attribute Table Exporter for analysis.
Use the probe alongside a colour-ramped display so the readouts confirm what your eye is interpreting.
Probe nodata buffer pixels around your area of interest to confirm the mask is working.
Switch coordinate format to MGRS or UTM if you are integrating field measurements into the workflow.
For multi-band rasters, name the bands in the probe panel so you remember which is red, NIR, etc.

Comparison with Other GIS Approaches

QGIS exposes raster query through the Identify Features tool and the Value Tool plugin. ArcGIS Pro provides the Pixel Inspector and the Identify tool. GDAL has gdallocationinfo for command-line scripts. The Raster Value Probe on gis.tools mirrors all of these in a zero-install browser interface. The advantage is that the entire raster is decoded once and held in memory, so probing is instant and there is no round-trip to a server. The trade-off is that very large rasters consume browser memory.

Performance Considerations

After the initial GeoTIFF decode (a one-time cost), every probe is an O(1) array lookup — effectively instantaneous. The raster is held in a typed array in browser memory; a 4000x4000 Float32 raster occupies 64 MB. Multi-gigabyte rasters need either downsampling with Raster Resampler or clipping with Clip Raster by Polygon.

Data Privacy and Browser-Based Processing

The raster file never leaves your machine. The File API reads the bytes, geotiff.js decodes them, and a typed array holds the pixels for the lifetime of the page. Probing does not generate network traffic. There is no upload, no telemetry, no logging. This matters for proprietary survey data, classified imagery and any compliance-sensitive workflow.

Related GIS Concepts

Cell Resolution: The pixel size of the raster, which determines spatial precision. A probe value at the cursor reflects the average over one cell.

Affine Transform: The six-number geotransform that maps pixel row/column to map coordinates. The probe relies on inverting this transform.

Nodata Value: A sentinel pixel value indicating "no measurement here". The probe explicitly reports nodata rather than misleading you with a numeric value.

Continuous vs Categorical Rasters: Continuous rasters (elevation, temperature, NDVI) tolerate interpolation; categorical rasters (land cover, soil class) must use nearest-neighbour sampling.

Bilinear vs Nearest Neighbour Sampling: The two main resampling strategies for reading sub-pixel values. See the Raster Resampler for a fuller treatment.

Frequently Asked Questions

Can I probe multiple bands at once?

Yes — for multi-band GeoTIFFs the probe returns the value of every band at the clicked location.

Does the probe work on land cover rasters?

Absolutely. Use nearest-neighbour sampling so the integer class codes are preserved. The probe will display the raw code; pair it with a legend lookup table to display the class name.

Why does my probe return 0 instead of the expected value?

Two common causes: the cell is masked nodata (and 0 is the nodata sentinel), or you are clicking outside the raster footprint. Confirm the bounds and the nodata value.

Can I export a list of probed values?

Yes — the gis.tools workflow lets you save the probe history and export it as CSV via the Attribute Table Exporter.

Does the probe support reprojection?

Yes. If your basemap is in EPSG:3857 and the raster is in UTM, the probe transforms coordinates on the fly.

How precise is the probe on touch devices?

Touch events register a roughly 10-pixel hit zone. Zoom in to ensure sub-pixel accuracy on small rasters.

How to Use