Field Delineation: Regrow’s ParcelID algorithm

What is Parcel ID?

Field boundaries are the foundation of everything Regrow measures. Before we can estimate a field's emissions, track a practice change, or establish a carbon baseline, we first need to know exactly where that field begins and ends.

Historically, many platforms relied on public or government data sets for this, such as the USDA's Common Land Units (CLUs) in the United States. Many of these are good, reliable datasets, but have some limitations when used for carbon program & sustainability use cases. For example, CLUs work reasonably well across the U.S. corn belt, but they have meaningful gaps: coverage is incomplete in many regions, and outside the core row-crop areas they frequently miss working agricultural land altogether. Parcels from European governments often have a delay in when they become available to the public.

To solve this, Regrow built its own field delineation product, called Parcel ID. Parcel ID delineates field boundaries directly from satellite imagery, giving consistent coverage across geographies.

The methodology

At a high level, Parcel ID looks at how a piece of land behaves over several years of satellite observations and uses those patterns to draw the lines between one field and the next.

Start with multi-year satellite imagery. Rather than looking at one image from one day, Parcel ID draws on several years of satellite imagery for each area (typically 5-7 prior years). Looking across many years lets the model see the persistent shape of a field, the footprint that shows up season after season, rather than a one-off pattern caused by a single crop, a wet spring, or a cloud.
Separate farmland from everything else. Not every green pixel in a remote sensing dataset is a working field. Before drawing any boundaries, Parcel ID masks out the land that isn't managed agriculture, such as roads, waterways, buildings, and natural non-cropped areas. It does this by combining public land-cover data with mapping data that identifies features like highways and rivers.
Identify distinct field boundaries. With farmland identified, the model examines the imagery and groups together the areas that consistently behave alike, and that are physically connected, into individual units. Where behavior changes, at a tree line, a fence row, a road, or the edge of a neighboring crop, the model places a boundary. The output at this stage is a first-pass map of every field in the area.
Clean up the edges. Raw boundaries drawn from satellite imagery can look jagged or pixelated. This is a normal and expected part of working with remote sensing data. Parcel ID applies a refinement step that smooths those edges into natural-looking field shapes while preserving the true size and form of each field, so a round center-pivot field comes out round and a rectangular field keeps its corners.
QA. Finally, the model reviews the boundaries it produced and removes ones that are not representative of true fields. For example, shapes that are too small to be a real field or that have unrealistic dimensions.

Why Parcels stay consistent over time

Customers who use Regrow to track change over time often ask a fair question: if a field's crops, tillage, and appearance change from season to season, how can the boundary around it stay consistent?

A field can be defined over different time horizons, and for different purposes. The agriculture industry recognizes several established boundary types, including:

Crop Sequence Boundary (CSB): delineates areas where different crops are rotated seasonally, reflecting current agricultural practices and the crop types in use.
Field Management History Boundary (FMHB): encompasses the historical record of agricultural practices within a field, including past crop rotations, tillage methods, and other management activities.
Ownership and Permanent Barrier Boundary (OPBB): aligns with legal property lines or permanent physical barriers such as fences or roads, defining the long-term, fixed extents of a farm or agricultural parcel.

Each of these is a valid, widely used way to represent a field; they simply answer different questions. Regrow's Parcel ID uses the Field Management History Boundary (FMHB) approach, because it's the best fit for understanding how a field has been managed, and measuring the outcomes of those practices, across multiple seasons.

Because the FMHB reflects a field's management history, Parcel ID looks across several years of satellite imagery rather than a single season. Over that window the crop growing in a field changes from year to year, but the outline of the managed area, defined by lasting features like field edges, hedgerows, farm roads, and drainage, stays largely the same. Parcel ID uses those persistent patterns to define the field, so the boundary represents the stable, managed footprint of the land rather than whatever happened to be planted in any one year. That's what allows a single, consistent boundary to represent a field across many seasons of measurement.

This doesn't mean boundaries are frozen forever. Managed areas genuinely do shift over time as fields are combined or split, or as new ground is brought into production, and the FMHB approach accounts for that by design. It captures the representative footprint across a field's recent management history rather than reacting to every seasonal change. The goal is a boundary that stays consistent enough to support reliable measurement over time, while still reflecting how the land is actually being managed.

ParcelID: USDA CLUs vs. Regrow Generated Field Boundaries

By using Regrow’s generated field boundaries, customers benefit from a higher level of precision compared to the less accurate legacy solutions provided by Common Land Units (CLUs), which lack comprehensive coverage, as detailed below.

Benefits of using Regrow generated field boundaries:

Regrow's Parcel ID uses satellite imagery to accurately delineate agricultural field boundaries within a region. This is crucial for generating SI data, as it provides a detailed list of fields, which are processed through Regrow’s Monitor and Measure APIs. Parcel ID simplifies MRV farmer enrollment by pre-filling field boundaries, making it easier for farmers to select their fields. It also improves the accuracy of covered area identification and the exclusion of non-agricultural land. Unlike outdated CLUs, which can complicate farmer pre-fill and credit auditing in MRV programs, Regrow's Parcel ID offers precise and current field delineations, enabling more efficient credit auditing.

In many regions outside the corn belt, we found CLUs often missed many agricultural fields, providing an incomplete picture of cropping practices.

Comparison of CLUs (left) and Regrow parcels in Missouri (right)

Comparison of CLUs (left) and Regrow parcels in Wisconsin (right)

Comparison of CLUs (left) and Regrow parcels in Massachusetts (right)