An overview of forest remote sensing technologies
HTML-код
- Опубликовано: 7 сен 2024
- Here I give a near complete overview of remote sensing technologies that can be used to measure forest ecosystems.
This one's very information-dense. So maybe take it slow and feel free to ask me questions!
The five technologies that I dive into here are:
1. Imaging
2. Radar
3. Photogrammetry
4. Hyperspectral
5. LiDAR
DISCLAIMER: I am not in any way affiliated with Pachama. I stopped working for them in Sept. 2021. The opinions and data in this video do not in any way reflect their own.
How does this video not have more views and likes, it is awesome!!! Informative, entertaining, great graphics and even humour! Thank you so much.
This is a great introductory overview of all things remote sensing for forests, not an easy thing to compile into one article!
I'll maybe comment on a few inaccuracies worth mentioning:
1) You missed the 1 to 5m imagery resolution. This is arguably the most useful resolution range for forest analysis and monitoring. Planet images nearly all of the Earth's forests every day at 3.7m and 8 spectral bands and studies have shown it is the most effective imagery data for detecting forest degradation. Note: you mention other lower resolution satellite image capture every day, but this is not the case, Landsat and Sentinel 2 combined for example do not provide more than 2 to 4 scenes per location per month.
2) Photogrammetry from spaceborne sensors is already here, not coming eventually as described. Numerous studies have shown using sub meter satellite imagery in 3d canopy modeling and inventories can produce very accurate volume and basal area estimates.
3) Hyperspectral is not the only way to estimate species and biodiversity, this is being done with high density lidar alone and even better with lidar and multispec imagery with very high accuracy. No hyperspectral required until you get more than about two dozen species in a stand.
4) Related to 3, time series imagery is very useful for species, ecosystem variables, and biodiversity estimation, and arguably more reliable than hyperspectral analysis due to signal redundancy
5) Some satellite based sub meter imagery is more more cost effective than drones and aerial imagery, enabling individual tree detection from space.
We are further along than you may suspect with satellite estimates of carbon. By generating models with satellite time series trained on lidar data we are getting very reliable AGB and carbon data at sub ha resolution, 3 to 5m specifically at national scales.
Thanks again for this overview on a rapidly evolving tech space!
This is wonderful, I worked with low-res satellite images for weather before, and this video provided a lot of new information to me. Thank you!!
great video. I love the quirkiness of your style. Gets the message straight out. Keep it up!
Fantastic video and super helpful, thanks!
That’s some great source of information. Very direct and vivid.
Can you please explain in brief about “Modelling Forest phenonological parameters using time series Remote Sensing Data.”
Thanks for going through this explaining! I loved the summary at the end 😃
Elias, you are a legend.
Indeed he is. All he needs is a bigger platform.
Hilarious and highly informative!
great video. i am using hyperspectral imaging for animal health.
love your content
great video! thank you so much!
Excelente!
What if Elias and Johnny Harris made a video about mapping the entire planet’s terrestrial forests? It would break the internet that’s what…
All theres are imagery. The firs point should be optical imagergy