Monitor API V3 technical migration guide

Use this guide to learn how to migrate from Monitor API V2 (current version) to V3 (new version).

We’re planning a new major version of Monitor API (v3), that comes with improved practice reporting and new features in the API. This release includes breaking changes to the API schema, that will require customers to update their integration in order to take advantage of the new features and continue using the API.

Use this guide for high level guidance on what's new and how to upgrade an existing integration. For more detailed integration information check out the full Integration Guide or the technical API reference.

What You'll Learn

After reading this guide, you'll understand the following:

New features and model improvements released in V3 of Monitor API
How to upgrade an integration with Monitor API V2 → V3
How to interpret changes to practice data

What's New in V3

Improved CropID ML Model for predicting crop type and growing season.

This release upgrades the CropID ML model (CropID V4), enhancing crop type prediction and introducing new functionality to predict planting and harvest dates using machine learning. These improvements provide more precise crop identification while enabling better tracking of field activity timelines.

Key Updates in the API:

Improved crop type classification for greater accuracy.
New planting and harvest date predictions, offering improved insights into field operations.
Additional crop types are supported in select regions. (CropID now supports lentils, chickpeas, flax, peanuts and durum wheat in CONUS and Canada)

CropID V4 will be released for CONUS and Canada in the Monitor V3 release. Refreshed data using CropID V4 in Europe will be coming soon!

Improved cover crop detection algorithm.

The this release includes an update to the cover crop detection algorithms to improve clarity, flexibility, and accuracy in vegetation coverage assessment. This version introduces simplified, monthly field-scale vegetation metrics based on NDVI, making it easier to track cover crop presence and vigor over time & apply insights to key use cases such as practice verification.

Key Updates in the API:

Simplified cover crop object for easier integration and more applicable use-case focused insights.
Cover crop determinations no longer reported on perennial fields for more precise monitoring.
Improved validation datasets enhance accuracy and reliability.

These changes provide a more interpretable & applicable method for tracking cover crop practice adoption, enabling better insights for tailored application.

Improved tillage classifications between three tillage classes: no till, reduced till, conventional till

With this release, we have improved the accuracy of tillage classification by refining residue percentage thresholds based on crop type. Previously, Monitor applied static USDA residue thresholds across all crops to distinguish between conventional, reduced-till, and no-till practices. While effective for high-biomass crops like corn, these thresholds proved less reliable for fragile crops—those with lower or rapidly degrading residue, such as cotton and soybeans.

The residue thresholds for fragile crops have been adjusted to improve the differentiation between reduced-till and no-till practices. This update reduces conservative bias and enhances alignment with observed tillage practices, particularly for crops with lower or rapidly degrading biomass that would change the expected residue persistence over time.

This release includes the update to the residue thresholds for fragile vs. non-fragile crops:

Note: Tillage thresholds are based on the crop that produced the residue, which is the crop in the previous cultivation cycle. For example, if a field grows corn one year and soybeans the next year, the residue & tillage practice for the soybean cycle will be based on the residue produced from the prior corn crop.

	Fragile crops	Non-fragile crops
Common crops in this category	Soybeans, winter wheat, cotton, cereal grains, potatoes, canola	Corn, rice, sorghum, sugarcane, hay/alfalfa
Conventional Till	0-15% residue	0-30% residue
Reduced till	15-30% residue	30-60% residue
No Till	30% + residue	60% + residue

New error codes added to make missing data scenarios more transparent and actionable.
This release includes three new error codes that provide more details and information about how to use and action different scenarios where remote sensing data may not be available. Codes `MD07`, `MD08` and `MD09` codes can now be returned from the API. Check out the Monitor Error Reporting Guide for additional details.

Migrating to V3 data schema

None of the existing Monitor API endpoints will change from V2 → V3, so no updates to the integration workflow are needed.

The field results schema is changing with the introduction of a new data points within the green cover object. This change requires an integration update to continue using the API.

Affected API Action(s):

API action	Endpoint	Description
Retrieve field practice data	/jobs/{job_id}/results (GET)	Use the jobID from the response of the execution post to retrieve the field events for the time period specified.

Cover crop data

Some of the legacy data points in the green cover object in V2 will be replaced with improved data points in Monitor V3.

Legacy green cover data point	New cover crop data point	Description
`start_date`	`start_date` no change	The first day of the cover crop observation window (the time period over which a field is observed for cover crop activity).
`end_date`	`end_date` no change	The last day of the cover crop observation window (the time period over which a field is observed for cover crop activity).
`cultivation_cycle_id`	`cultivation_cycle_id` no change	The unique identifier of the cultivation cycle the cover crop practice belongs to.
`green_cover_class`	`cover_crop_class`	The cover_crop_class is the practice determination over the observation period. Cover crop class options include: "Cover crop" "Cover crop not tracked" "Potential cover crop" "Cover crop not applicable" - Note: this is a new cover crop class that will be applied when a perennial crop is detected on the field "No green cover data"
`emergence_quality`	`emergence_class`	emergence_class is a classification of how a crop emerged in the first three months of the cover crop observation period. Emergence can be used to understand if a cover crop may have been planted, and if it established strongly. Possible options include: "No vegetation" "Weak emergence" "Good emergence" "Strong emergence"
`emergence_quality_confidence`	`emergence_class_confidence`	A confidence metric associated with the emergence_class. Confidence is reported as a scale between 1-3, where values correspond to low, medium and high confidence. Emergence confidence is an indicator of the amount of high-quality, interpretable remote sensing data available over the observation period.
`living_root quality_score`	`ndvi_mean`	The NDVI mean is the average monthly NDVI over the cover crop observation period. It can be used to evaluate the average greenness of the crop, providing an indication of crop strength and vigor over the growing period. ndvi_mean is reported on a scale of 0-100.
`overall_ confidence`	`cover_crop_class_confidence`	cover_crop_class_confidence is an indicator of confidence in the cover crop classification. Confidence is reported as a scale between 1-3, where values correspond to low, medium and high confidence. Cover crop class confidence is an indicator of the amount of high-quality, interpretable remote sensing data available over the observation period.
`living_root_score`	Not applicable deprecated in V3
`valid_observation_percent`	Not applicable deprecated in V3
`healthy_vegetation_coverage_percent`	Not applicable deprecated in V3
`vegetation_vigor`	Not applicable deprecated in V3

Example of the new green cover object:

Tillage data

The tillage object schema within field results is not changing. There are no breaking changes to API structure for tillage —only improved classification logic.

What this means for using Monitor tillage determinations:

Improved classification accuracy for fragile crops, reducing misidentification of tillage practices.
Potential shifts in classification outputs for specific crop types, leading to more reliable downstream insights.

To ensure seamless migration, review any tillage-dependent workflows or models that may be impacted by updated classifications.