|
''This page is about the intersection of forestry and machine learning in the context of climate change mitigation. For an overview of land use as a whole, please see the [https://en.wikipedia.org/wiki/Land_use Wikipedia page] on this topic.''[[File:Agriculture.png|thumb|A schematic of the ways that machine learning can support carbon negative agriculture, forestry, and land use.]]
AsThe described in the paper "Tackling Climate Change with Machine Learning"<ref>{{Cite journal|last=Rolnick|first=David|last2=Donti|first2=Priya L.|last3=Kaack|first3=Lynn H.|last4=Kochanski|first4=Kelly|last5=Lacoste|first5=Alexandre|last6=Sankaran|first6=Kris|last7=Ross|first7=Andrew Slavin|last8=Milojevic-Dupont|first8=Nikola|last9=Jaques|first9=Natasha|last10=Waldman-Brown|first10=Anna|last11=Luccioni|first11=Alexandra|date=2019-11-05|title=Tackling Climate Change with Machine Learning|url=http://arxiv.org/abs/1906.05433|journal=arXiv:1906.05433 [cs, stat]}}</ref>,<blockquote>Plants, microbes, and other organisms have been drawing CO2 from the atmosphere for millionsdeterioration of years. Most of this carbon is continually broken down and recirculated through the carbonnatural cycle, and someworld is storedunparalleled deepin undergroundhuman as coalhistory and oil, but a largekey amountdriver of carbon is sequestered in the biomassclimate of trees,crisis. peatSince bogs2000, andwe soil.have Ourlost current361 economymillion encourages practices that are freeing muchha of thisforest sequesteredcover carbon(roughly throughthe deforestation and unsustainable agriculture. On topsize of these effects, cattle and rice farming generate methane, a greenhouse gas far more potent than CO2 itself. Overall, land use by humans is estimated to beEurope) responsibleaccounting for about a quarter of global GHGanthropogenic emissions<ref>{{Cite book|title=Climate Change 2014: Mitigation of Climate Change. Contribution of Working Group III to the Fifth Assessment Report of the Intergovernmental Panel on Climate Change|url=https://www.ipcc.ch/report/ar5/wg3/|date=2014|coeditors=|last=|first=|publisher=|year=|isbn=|location=|pages=}}</ref> (and this may be an underestimate<ref>{{Cite journal|title=Are the impacts of land use on warming underestimated in climate policy?|url=https://iopscience.iop.org/article/10.1088/1748-9326/aa836d|coauthors=|date=2017|last=Mahowald|first=Natalie|journal=|volume=|pages=|via=}}</ref>)., Inlargely additiondriven toby thisdeforestation directand releaseforest ofdegradation. Deforestation does not only release carbon (e.g., through slash-and-burn), humanbut actionsalso destroys a multitude of other forest ecosystem services: preserving biodiversity, counteracting flooding and soil erosion, filtering water, and offering a livelihood for the permafrostlocal population. Major conservation and restoration efforts are underway to mitigate and safeguard against these losses, and to highlight the urgency of the issue, 2021-2030 has been declared the “UN Decade on Ecosystem Restoration”. However, we cannot preserve what we cannot measure. There is nowan meltingurgent need to understand the location, peathealth bogsand areecological value of nature and dryingbiodiversity, and forestensure firesthese metrics are becomingreflected morein frequentpolicy, asfinance, and decision-making. Machine learning (ML) can play a consequencesignificant ofrole climatein changeresponding itselfto –this allcritical ofcall whichfor releaseaction yetand morecan carbon<ref>{{Citeaccelerate journal|title=Thethe studyconservation and sustainable use of Earthforestry asand another integratedland system|url=https://climateuse.nasa.gov/nasa_science/science/# [[File:~:text=The%20Study%20of%20Earth%20as%20an%20Integrated%20System&text=Earth%20system%20science%20is%20theAgriculture.png|thumb|A schematic of the ways that machine learning can support carbon negative agriculture,whole%2C%20including%20its%20changing%20climate.}}</ref> forestry, and land use.]]
== Machine Learning Application Areas == ▼The large scale of this problem allows for a similar scale of positive impact. According to one estimate<ref>{{Cite book|title=Drawdown: The most comprehensive plan ever proposed to reverse global warming|author=Paul Hawken|date=2015}}</ref>, about a third of GHG emissions reductions could come from better land management and agriculture. ML can play an important role in somehelping ofus thesetackle areasclimate change through land use. Precision agriculture could reduce carbon release from the soil and improve crop yield, which in turn could reduce the need for deforestation. Satellite images make it possible to estimate the amount of carbon sequestered in a given area of land, as well as track GHG emissions from it. ML can help monitor the health of forests and peatlands, predict the risk of fire, and contribute to sustainable forestry. These areas represent highly impactful applications, in particular, of sophisticated computer vision tools, though care must be taken in some cases to avoid negative consequences via the Jevons paradox. </blockquote>▼
=== Ecosystem monitoring ===
▲The large scale of this problem allows for a similar scale of positive impact. According to one estimate<ref>{{Cite book|title=Drawdown: The most comprehensive plan ever proposed to reverse global warming|author=Paul Hawken|date=2015}}</ref>, about a third of GHG emissions reductions could come from better land management and agriculture. ML can play an important role in some of these areas. Precision agriculture could reduce carbon release from the soil and improve crop yield, which in turn could reduce the need for deforestation. Satellite images make it possible to estimate the amount of carbon sequestered in a given area of land, as well as track GHG emissions from it. ML can help monitor the health of forests and peatlands, predict the risk of fire, and contribute to sustainable forestry. These areas represent highly impactful applications, in particular, of sophisticated computer vision tools, though care must be taken in some cases to avoid negative consequences via the Jevons paradox.</blockquote>
''Main article: [[Ecosystem Monitoring]]''
▲== Machine Learning Application Areas ==
Monitoring the nature's health is a key element in policies surrounding its protection. Advancements in machine learning and remote sensing allow us to scale measurement, reporting and verification processes at an unprecedented scale.
=== Monitoring ecosystem restoration ===
* Biodiversity monitoring
* Estimating carbon stock ▼* Peatland monitoring
* Forest health monitoring
*Reducing deforestation
* Blue carbon monitoring
=== '''ManagingLand forestsuse management''' ===
Main article: ''[[Land Use Management]]''
* Automating afforestation
* ManagingForecasting forest fires
* Reducing deforestation
*Empowering forest communities
*Supply chain optimisation
▲* Estimating carbon stock
== Background Readings ==
=== OtherKey documents ===
[https://www.ipcc.ch/srccl/ Special Report on Climate Change and Land — IPCC site]
*'''"Monitoring tropical forest carbon stocks and emissions using Planet satellite data" (2019)'''<ref>{{Cite journal|last=Csillik|first=Ovidiu|last2=Kumar|first2=Pramukta|last3=Mascaro|first3=Joseph|last4=O’Shea|first4=Tara|last5=Asner|first5=Gregory P.|date=2019-11-28|title=Monitoring tropical forest carbon stocks and emissions using Planet satellite data|url=http://dx.doi.org/10.1038/s41598-019-54386-6|journal=Scientific Reports|volume=9|issue=1|doi=10.1038/s41598-019-54386-6|issn=2045-2322}}</ref>: An example of carbon stock estimation from satellite imagery.
*'''"Characterizing agricultural drought in the Karamoja subregion of Uganda with meteorological and satellite-based indices" (2018)'''<ref>{{Cite journal|last=Nakalembe|first=Catherine|date=2018-02-10|title=Characterizing agricultural drought in the Karamoja subregion of Uganda with meteorological and satellite-based indices|url=http://dx.doi.org/10.1007/s11069-017-3106-x|journal=Natural Hazards|volume=91|issue=3|pages=837–862|doi=10.1007/s11069-017-3106-x|issn=0921-030X}}</ref>: A study of drought phenomena using remote sensing data.
*'''"Remote Sensing Northern Lake Methane Ebullition" (2020)'''<ref>{{Cite journal|last=Engram|first=M.|last2=Walter Anthony|first2=K. M.|last3=Sachs|first3=T.|last4=Kohnert|first4=K.|last5=Serafimovich|first5=A.|last6=Grosse|first6=G.|last7=Meyer|first7=F. J.|date=2020-05-11|title=Remote sensing northern lake methane ebullition|url=http://dx.doi.org/10.1038/s41558-020-0762-8|journal=Nature Climate Change|volume=10|issue=6|pages=511–517|doi=10.1038/s41558-020-0762-8|issn=1758-678X}}</ref>: A study relating methane emissions estimates from airplanes and satellite imagery.
== Online Courses and Course Materials ==
[https://www.nationalgeographic.org/projects/exploring-conservation/ National Geographic Free Conservation Course]
[https://www.coursera.org/learn/spatial-analysis UC Davis Geospatial and Environmental Analysis]
[https://www.coursera.org/learn/ecosystem-services University of Geneva's Introduction to Ecosystem Services]
== Conferences, Journals, and Professional Organizations ==
== Community ==
=== Major conferences ===
* [https://www.iucncongress2020.org/ IUCN World Congress] (every four years)
*'''EARTHVISION''': A workshop regularly held at computer vision conferences. Website [https://www.grss-ieee.org/earthvision2020/ here].
* [http://www.fao.org/about/meetings/cofo/en/ Committee on Forestry] (annual)
*'''Space and AI:''' A conference organized by the ESA-CLAIRE AI Special Interest Group on Space. Website [https://claire-ai.org/sig-space/?lang=fr here].
* [https://www.agu.org/fall-meeting AGU Fall Meeting] (annual)
* [https://naturalcapitalproject.stanford.edu/events/2019-natural-capital-symposium Natural Capital Symposium] (annual)
* [https://www.globallandscapesforum.org/ Global Landscapes Forum] (annual)
=== Major journals ===
* [https://www.nature.com/ Nature] (Climate Change, Ecology and Evolution)
* [https://www.sciencemag.org/ Science]
* [https://www.journals.elsevier.com/remote-sensing-of-environment Remote Sensing of Environment]
* [https://www.mdpi.com/journal/remotesensing Remote Sensing]
=== Major societies and organizations ===
* [http://www.fao.org/home/en/ The Food and Agriculture Organization of the United Nations] (FAO) is a specialized agency of the United Nations that leads international efforts to defeat hunger and improve nutrition and food security
* [https://www.thegef.org/ The Global Environment Facility] (GEF) was established on the eve of the 1992 Rio Earth Summit to help tackle our planet’s most pressing environmental problems.
* [https://www.nature.org/en-us/ The Nature Conservancy] is a non-profit environmental organisation that has over one million members, and has protected more than 119,000,000 acres (48,000,000 ha) of land and thousands of miles of rivers worldwide
* [https://www.iucn.org/ The International Union for Conservation of Nature] (IUCN) is an international organization working in the field of nature conservation and sustainable use of natural resources.
*[https://www.southpole.com/ SouthPole], develops and finance projects around the world to reduce carbon emissions and protect biodiversity
*[https://www.goldstandard.org/ GoldStandard], is a third-party certification agency that ensures projects that reduced carbon emissions featured the highest levels of environmental integrity and also contributed to sustainable development
*[https://www.worldwildlife.org/ WWF]
*[https://www.worldbank.org/ World Bank]
*[https://www.1t.org/ 1t.org]
*[http://www.mangrovealliance.org/ Global Mangrove Alliance]
=== Major initiatives ===
* [https://www.unenvironment.org/news-and-stories/press-release/new-un-decade-ecosystem-restoration-offers-unparalleled-opportunity UN Decade of Ecosystem Restoration]
== Libraries and Tools ==
Some packages forand workingtools withthat remotesupport sensingmachine datalearning-based forestry work are,
*[https://restor.eco Restor.eco], an open data platform for the global restoration movement
*'''eo-learn''': A python package maintained by the European Space Agency, giving easy access to imagery from Sentinel satellites, as well as utilities for data processing, available [https://github.com/sentinel-hub/eo-learn here].
*[https://www.globalforestwatch.org/ GlobalForestWatch], is a dynamic online forest monitoring and alert system that empowers people everywhere to better manage forests
*'''robosat''': A package maintained by mapbox, available [https://github.com/mapbox/robosat here].
*[https://www.globalmangrovewatch.org/ GlobalMangroveWatch], monitors to catalyse the action needed to protect and restore mangroves
*'''solaris''': A package from CosmiQ Works (SpaceNet Challenge), available [https://solaris.readthedocs.io/en/latest/ here].
*[https://trase.earth/ Trase.Earth] provides a dynamic supply chain visualisations for land change drivers and commodities
*[https://resourcetrade.earth/ Resourcetrade.Earth] features Chatham House’s extensive and authoritative database of international trade in natural resources, developed from United Nations data
*[http://www.openforis.org/tools/collect-earth.html Collect Earth] is a free, open source, and user-friendly tool using Google Earth Engine to visualize and analyze plots of land in order to assess deforestation and other forms of land-use-change
*[https://www.wri.org/our-work/project/forest-atlases/open-data-portals The Forest Atlases] are online platforms that help countries better manage their forest resources by combining government data with the latest forest monitoring technology
*[https://github.com/weecology/DeepForest DeepForest], is a python package for tree crown detection in airborne RGB imagery
*[https://github.com/ulaval-damas/tree-bark-classification BarkNet], is an open-source tree bark classification algorithm and dataset
== Data ==
* [https://www.kaggle.com/c/planet-understanding-the-amazon-from-space Planet: Understanding the Amazon from space]
* [https://idtrees.org/ IDTReeS: Integrating Data science with Trees and Remote Sensing]
Tree datasets:
* [hhttps://github.com/selva-lab-repo/TALLO TALLO: A global tree allometry and crown architecture database]
* [https://github.com/blutjens/awesome-forests Awesome-forests is a curated list of ground-truth/validation/in situ forest datasets for the forest-interested machine learning community]
== References ==
|