Skip to main content
Solved

A Question for the Community: Is Tree-Level Monitoring from Space Possible?

  • July 8, 2026
  • 1 reply
  • 17 views

Mario WWW
Forum|alt.badge.img+1

I'm currently developing Canopia Agrotech as part of my Master's thesis in Innovation and Entrepreneurship, and I'm exploring the next generation of satellite imagery for precision agriculture.

The goal of the project is to monitor individual trees in woody crops (such as olive, almond, and pistachio orchards), detect early signs of stress or disease, and provide actionable insights that help farmers reduce losses and optimize field management.

Most current agricultural platforms rely on multispectral imagery (e.g., Sentinel-2), which works well at the parcel level. However, my research is investigating whether combining higher spatial resolution, hyperspectral imagery, and AI could make reliable tree-level monitoring possible from satellite data.

I'm curious:

  • Does Planet currently offer hyperspectral imagery, or have plans to integrate it into its commercial products?
  • Have you seen successful applications of hyperspectral satellite data for tree-level health monitoring in agriculture?
  • Would you recommend combining Planet imagery with other hyperspectral data sources for this type of use case?

I'd love to hear your thoughts or be pointed toward any relevant resources or case studies.

Thank you!
#PrecisionAgriculture #RemoteSensing #Hyperspectral #EarthObservation #SatelliteImagery #AI #AgTech #MastersThesis

Best answer by christopher

Hi ​@Mario WWW , great questions!

Planet does operate a satellite hyperspectral instrument — Tanager — but it is a 30 m resolution tasking system. Tree-level hyperspectral analyses often use airborne imaging spectrometers to get tree-scale data (1-3 m resolution). You can check out this special issue dedicated to tree-level data science for tree mapping using high res airborne data, for example.

Here are some useful resources for addressing your research questions, albeit indirectly:

PlanetScope provides 8-band multispectral data that are great for monitoring vegetation health, including red edge and near infrared channels. I recommend working with normalized surface reflectance mosaics or analysis-ready planetscope (ARPS) to improve normalization and consistency.

To find how others are using PlanetScope data in practice, you can search for relevant use cases in the Publications Database.

You might also be interested our forest monitoring products like Forest Carbon Monitoring. This includes 3m resolution estimates of canopy height and canopy cover that are updated quarterly, which are effective for mapping tree extent and tree change over time (though these are not designed to evaluate tree health).

We recently provided two open data releases related to agroforestry: the Individual Tree Detection benchmark dataset and the Agroforestry Tree Species Identification dataset. You might find these useful for R&D.

While high resolution hyperspectal is not available through Planet we do provide a great series of tools for high resolution tree mapping and tree monitoring that I hope you find useful. There’s no “one” way to evaluate tree stress or tree health, so it is hard to build a single product that does so, but these tools should provide the ingredients necessary for you to develop a monitoring solution that works in your context.

Good luck!

1 reply

christopher
Planeteer 🌎
Forum|alt.badge.img+2
  • Planeteer 🌎
  • Answer
  • July 9, 2026

Hi ​@Mario WWW , great questions!

Planet does operate a satellite hyperspectral instrument — Tanager — but it is a 30 m resolution tasking system. Tree-level hyperspectral analyses often use airborne imaging spectrometers to get tree-scale data (1-3 m resolution). You can check out this special issue dedicated to tree-level data science for tree mapping using high res airborne data, for example.

Here are some useful resources for addressing your research questions, albeit indirectly:

PlanetScope provides 8-band multispectral data that are great for monitoring vegetation health, including red edge and near infrared channels. I recommend working with normalized surface reflectance mosaics or analysis-ready planetscope (ARPS) to improve normalization and consistency.

To find how others are using PlanetScope data in practice, you can search for relevant use cases in the Publications Database.

You might also be interested our forest monitoring products like Forest Carbon Monitoring. This includes 3m resolution estimates of canopy height and canopy cover that are updated quarterly, which are effective for mapping tree extent and tree change over time (though these are not designed to evaluate tree health).

We recently provided two open data releases related to agroforestry: the Individual Tree Detection benchmark dataset and the Agroforestry Tree Species Identification dataset. You might find these useful for R&D.

While high resolution hyperspectal is not available through Planet we do provide a great series of tools for high resolution tree mapping and tree monitoring that I hope you find useful. There’s no “one” way to evaluate tree stress or tree health, so it is hard to build a single product that does so, but these tools should provide the ingredients necessary for you to develop a monitoring solution that works in your context.

Good luck!