Hybrid regression method to predict forest variables from Earth observation data in boreal forests
Satellite remote sensing is essential for monitoring the boreal forest, the largest land biome on Earth. With the growing volume of Earth observation (EO) data and increasing demand for actionable information, more efficient and robust monitoring methods are needed. Machine learning-based approaches...
Saved in:
| Main Authors: | , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Taylor & Francis Group
2025-12-01
|
| Series: | European Journal of Remote Sensing |
| Subjects: | |
| Online Access: | https://www.tandfonline.com/doi/10.1080/22797254.2025.2462032 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1849687571257360384 |
|---|---|
| author | Eelis Halme Matti Mõttus |
| author_facet | Eelis Halme Matti Mõttus |
| author_sort | Eelis Halme |
| collection | DOAJ |
| description | Satellite remote sensing is essential for monitoring the boreal forest, the largest land biome on Earth. With the growing volume of Earth observation (EO) data and increasing demand for actionable information, more efficient and robust monitoring methods are needed. Machine learning-based approaches offer flexibility but rely on extensive training data, which can be generated with reflectance models. This study introduces a hybrid regression method, integrating the forest reflectance and transmittance model FRT with a random forest regressor. Using a representative dataset from Finland (24 081 plots), the method was trained to predict structural boreal forest variables: mean height, mean diameter at breast height (DBH) and basal area from EO data. The prediction performance was evaluated using three independent test areas, two from Finland and one from Sweden. In Finland, the most accurate predictions had root-mean-square errors of 3.6 m (19.1%) for height, 6.3 cm (27.3%) for DBH and 9.9 m2 ha−1 (31.6%) for basal area. In Sweden, low R2 values (< 0.1) indicated limitations in transferability. The results suggest that combining reflectance modelling with machine learning can advance environmental monitoring methodologies in the boreal forest but also demonstrate the challenges of applying these methods across different geographical regions. |
| format | Article |
| id | doaj-art-0877f2f882f44db9ba830b09067e20c7 |
| institution | DOAJ |
| issn | 2279-7254 |
| language | English |
| publishDate | 2025-12-01 |
| publisher | Taylor & Francis Group |
| record_format | Article |
| series | European Journal of Remote Sensing |
| spelling | doaj-art-0877f2f882f44db9ba830b09067e20c72025-08-20T03:22:18ZengTaylor & Francis GroupEuropean Journal of Remote Sensing2279-72542025-12-0158110.1080/22797254.2025.2462032Hybrid regression method to predict forest variables from Earth observation data in boreal forestsEelis Halme0Matti Mõttus1VTT Technical Research Centre of Finland, Espoo, FinlandVTT Technical Research Centre of Finland, Espoo, FinlandSatellite remote sensing is essential for monitoring the boreal forest, the largest land biome on Earth. With the growing volume of Earth observation (EO) data and increasing demand for actionable information, more efficient and robust monitoring methods are needed. Machine learning-based approaches offer flexibility but rely on extensive training data, which can be generated with reflectance models. This study introduces a hybrid regression method, integrating the forest reflectance and transmittance model FRT with a random forest regressor. Using a representative dataset from Finland (24 081 plots), the method was trained to predict structural boreal forest variables: mean height, mean diameter at breast height (DBH) and basal area from EO data. The prediction performance was evaluated using three independent test areas, two from Finland and one from Sweden. In Finland, the most accurate predictions had root-mean-square errors of 3.6 m (19.1%) for height, 6.3 cm (27.3%) for DBH and 9.9 m2 ha−1 (31.6%) for basal area. In Sweden, low R2 values (< 0.1) indicated limitations in transferability. The results suggest that combining reflectance modelling with machine learning can advance environmental monitoring methodologies in the boreal forest but also demonstrate the challenges of applying these methods across different geographical regions.https://www.tandfonline.com/doi/10.1080/22797254.2025.2462032Sentinel-2boreal foresthybrid inversionmachine learningreflectance modelSwedish NFI |
| spellingShingle | Eelis Halme Matti Mõttus Hybrid regression method to predict forest variables from Earth observation data in boreal forests European Journal of Remote Sensing Sentinel-2 boreal forest hybrid inversion machine learning reflectance model Swedish NFI |
| title | Hybrid regression method to predict forest variables from Earth observation data in boreal forests |
| title_full | Hybrid regression method to predict forest variables from Earth observation data in boreal forests |
| title_fullStr | Hybrid regression method to predict forest variables from Earth observation data in boreal forests |
| title_full_unstemmed | Hybrid regression method to predict forest variables from Earth observation data in boreal forests |
| title_short | Hybrid regression method to predict forest variables from Earth observation data in boreal forests |
| title_sort | hybrid regression method to predict forest variables from earth observation data in boreal forests |
| topic | Sentinel-2 boreal forest hybrid inversion machine learning reflectance model Swedish NFI |
| url | https://www.tandfonline.com/doi/10.1080/22797254.2025.2462032 |
| work_keys_str_mv | AT eelishalme hybridregressionmethodtopredictforestvariablesfromearthobservationdatainborealforests AT mattimottus hybridregressionmethodtopredictforestvariablesfromearthobservationdatainborealforests |