Understanding your Vegetation Disturbance Report
The disturbance monitoring system uses imagery from two families of Earth observation satellites: Landsat (operated by NASA and the US Geological Survey) and Sentinel-2 (operated by the European Space Agency). Both orbit the Earth at an altitude of roughly 700 km and carry cameras capable of capturing detailed images of the land surface at a resolution of 30 metres per pixel — meaning each individual pixel in the image represents a 30 × 30 metre patch of ground.
As a satellite moves along its orbit it does not photograph one point at a time — its camera captures a wide strip of land stretching from side to side. This strip is called a swath. For both Landsat and Sentinel-2 the swath is approximately 180–290 km wide. On each orbit the satellite records a different swath, and over the course of about 10–16 days it has covered the entire surface of the Earth. The swaths are divided into standardised tile grids so that imagery from different satellites and different dates always lines up perfectly.
Because Landsat and Sentinel-2 are on different orbital schedules, together they can image any given location on Earth roughly every 2–3 days (weather permitting), compared with roughly every 8–10 days for either satellite family on its own. This combined view is what makes near real-time disturbance monitoring possible.
A time series is simply a sequence of observations of the same place, made at regular intervals over time. In the context of this report, it means the system regularly checks the same site using the latest available satellite images and compares what it sees now with what it saw before.
Over the monitoring period you defined, dozens or even hundreds of individual satellite images may have been taken of your site. Rather than presenting every single image, the system identifies the key moments when the vegetation status changed — for example, when disturbance was first detected, when it was confirmed, and how it evolved over subsequent weeks and months.
The date previews in the report form a time series. You can step through them using the Prev / Next navigation buttons to see how disturbance appeared and spread across your site over time — similar to flipping through frames of a film.
One important constraint is cloud cover. Satellites capture reflected sunlight, so thick cloud blocks the view of the ground entirely. On cloudy days no useful observation is possible, which is why some periods may appear to have fewer observations than others.
DIST-ALERT is a data product produced by NASA as part of the OPERA (Observational Products for End-Users from Remote Sensing Analysis) programme. It analyses every new satellite image as it arrives and compares each pixel with the historical record of that same location going back to 2017. When a pixel looks significantly different from its historical behaviour — less green, less vegetated — the system flags it as a potential disturbance.
Because a single unusual observation might just be clouds, shadows, or a sensor artefact, DIST-ALERT uses a confidence ladder. A disturbance starts at low confidence and is only promoted as more observations confirm it:
Each confirmed pixel is also classified by severity — whether the vegetation loss is less than 50% or 50% or greater. A pixel with ≥50% loss represents a more severe disturbance, such as clear-felling or intense fire damage.
Once a pixel reaches the Confirmed state it stays confirmed — it does not reset. This means that the most recent DIST-ALERT image for any location shows the cumulative picture of all disturbance that has ever been detected there, not just what happened recently. A pixel confirmed in January will still show as confirmed in December of the same year, even if vegetation has begun to regrow. This cumulative nature is by design: it ensures that past disturbance events are not forgotten as the satellite continues to collect data.
Data availability lag
OPERA DIST-ALERT tiles are typically available 2–14 days after the satellite pass due to processing and ingest time. Sentinel-2 revisits a given location roughly every 5 days, and tiles may take a further 2–7 days to appear in the archive. This means a weekly report run on Monday may not yet include detections from the most recent few days of the monitoring window. The Monitoring Period shown in your report reflects the actual date range of granules found, not the originally requested window.
Because DIST-ALERT is cumulative, simply counting the confirmed pixels in the latest image would overcount — it would include disturbances that happened long before your monitoring period started. The system therefore uses a comparison approach: it looks at the state of vegetation at the very start of your monitoring period and compares it with the state at the very end.
The key question: "Which pixels were undisturbed (or only provisionally disturbed) at the start of the monitoring period, but had reached confirmed disturbance by the end?"
Two satellite granules are compared for each site:
For every 30 × 30 metre pixel within your site boundary, the system applies three rules — new disturbance, confirmed this period, and re-disturbed — which are described in detail, with the exact class values used, in the Cumulative Disturbance Map section below. Pixels that were already confirmed before your monitoring period started are excluded; they represent pre-existing damage, not something that happened on your watch.
The disturbed area in hectares and the percentage of the site affected are then calculated simply by counting the flagged pixels and multiplying by the area each pixel covers (900 m² for a 30 × 30 m pixel = 0.09 ha per pixel).
The cumulative disturbance map (CDM) is the red-shaded overview image shown at the top of your report. It marks every 30 × 30 m pixel within your site that was flagged as newly disturbed during this monitoring period. Understanding exactly which pixels qualify requires knowing both the DIST-ALERT class values and the comparison rules applied between the first and last granules.
Every pixel in a DIST-ALERT granule carries one of nine integer class values describing its current state. Classes 1–8 represent some form of detected disturbance; class 0 means no disturbance has been detected.
| Value | Class name | Confidence | Severity | Active? |
|---|---|---|---|---|
| 0 | No Disturbance | — | — | — |
| 1 | First Detection (<50% Loss) | Initial | Low | Yes |
| 2 | Provisional (<50% Loss) | Provisional | Low | Yes |
| 3 | Confirmed (<50% Loss) | Confirmed | Low | Yes |
| 4 | First Detection (≥50% Loss) | Initial | High | Yes |
| 5 | Provisional (≥50% Loss) | Provisional | High | Yes |
| 6 | Confirmed (≥50% Loss) | Confirmed | High | Yes |
| 7 | Confirmed (<50% Loss) Finished | Finished | Low | No |
| 8 | Confirmed (≥50% Loss) Finished | Finished | High | No |
For each pixel the system compares its class in the baseline granule (first granule of the monitoring period) with its class in the current granule (last granule). A pixel is shaded red on the CDM if it satisfies any one of the three rules below:
| Rule | Baseline class (first granule) | Current class (last granule) | Interpretation |
|---|---|---|---|
| New disturbance | 0 or absent | 3, 6, 7, or 8 | Pixel was undisturbed (or not present) at the start and is now confirmed or finished. Entirely new damage within this period. |
| Confirmed this period | 1, 2, 4, or 5 | 3, 6, 7, or 8 | Pixel was already showing early or provisional signs at the start and crossed the confirmed threshold during this period. |
| Re-disturbed | 7 or 8 | 3 or 6 | Pixel had previously finished disturbance at the start but has become actively confirmed again — a second disturbance event at the same location. |
The date previews section of the report shows a visual record of your site at key moments during the monitoring period. Each preview consists of two side-by-side images taken on the same date.
True colour image
The left-hand image shows the site as it would appear to the human eye — green for healthy vegetation, brown for bare soil, and so on. This is the raw satellite photograph.
Disturbance overlay
The right-hand image shows the same photograph with the DIST-ALERT disturbance data overlaid in red. Red pixels indicate areas of confirmed vegetation loss within your site boundary.
Three types of date are included:
Each date preview has ← Prev and Next → navigation buttons positioned above the image card. Clicking these jumps you to the previous or next date preview in the sequence. If you click repeatedly on Next → you can step through the entire monitoring period like a flipbook, watching disturbance appear and spread across the site over time.
Below the overlay image on each date preview there is a small summary table showing how many pixels of each disturbance class were present in the DIST-ALERT data on that date, and what area they cover. Note that this table reflects the raw DIST-ALERT classification for that moment in time — it includes all disturbance classes visible, not only the ones that fall within your monitoring period window. The headline disturbance statistics at the top of the report, by contrast, reflect only the disturbance that was newly confirmed during your monitoring period.
Pre-existing confirmed pixels: Any pixel that was already at confirmed status at the very start of your monitoring period will appear in the raw disturbance table but is excluded from all headline statistics and the cumulative disturbance map. It represents damage that predates this monitoring window and is not counted as new change. If you see confirmed pixels in the baseline date preview, you can safely ignore them — they are not contributing to the figures in your report.
© 2025 Restor Monitoring Services · OPERA DIST-ALERT data courtesy of NASA / JPL