Customer Feedback - Carbon Diligence Product

Related products: Planetary Variables

Feedback from a customer (conversation to be continued through further discussions):

“After evaluating the above-ground biomass product at 30m spatial resolution from Planet to detect forest harvests, we concluded that although it has a high harvest detection rate, it also has a high amount of noise (pixels classified as harvest in areas that were not harvested) (context: we classified pixels as harvest vs non-harvests by setting a threshold based on biomass change). Therefore, we decided to move forward with a different product to accomplish this task. However, we’ll keep evaluating the best products in the market, including Planet’s, on an annual basis to make sure we have the most suitable product for our purpose. I hope to touch base with you again later this year and I’m sure by then you’ll have an even better product!”

 

@david.marvin 

Thanks for the feedback! There are a couple of updates worth sharing.

  • First is v1.1 product update. This version update was specifically designed to reduce year-over-year variation in Canopy Height and Aboveground Carbon Density (see the Release Notes for more details). We believe this should significantly reduce the false positive harvest rate these users reported.
  • An alternative approach to estimating harvest/non-harvest area would be to use the Canopy Cover data instead of the Aboveground Carbon Density data. Canopy Cover is directly sensitive to harvest patterns, and it can be easier to set change thresholds based on easy-to-interpret numbers. Canopy Cover data are provided in units of % cover, and you could estimate harvests using a change threshold where >30% canopy cover changed occurred, for example)

We appreciate this feedback and our team will continue to work on solving these kinds of problems on behalf of our users.