Scientific publications

@article{nokey,

title = {Building a community-based open harmonised reference data repository for global crop mapping},

author = {Hendrik Boogaard and Arun Kumar Pratihast and ,Juan Carlos Laso Bayas, and Santosh Karanam, and Steffen Fritz, and Kristof Van Tricht, and Jeroen Degerickx, and Sven Gilliams},

url = {https://journals.plos.org/plosone/article?id=10.1371/journal.pone.0287731},

doi = {10.1371/journal.pone.0287731},

year  = {2023},

date = {2023-07-13},

urldate = {2023-07-13},

abstract = {Reference data is key to produce reliable crop type and cropland maps. Although research projects, national and international programs as well as local initiatives constantly gather crop related reference data, finding, collecting, and harmonizing data from different sources is a challenging task. Furthermore, ethical, legal, and consent-related restrictions associated with data sharing represent a common dilemma faced by international research projects. We address these dilemmas by building a community-based, open, harmonised reference data repository at global extent, ready for model training or product validation. Our repository contains data from different sources such as the Group on Earth Observations Global Agricultural Monitoring Initiative (GEOGLAM) Joint Experiment for Crop Assessment and Monitoring (JECAM) sites, the Radiant MLHub, the Future Harvest (CGIAR) centers, the National Aeronautics and Space Administration Food Security and Agriculture Program (NASA Harvest), the International Institute for Applied Systems Analysis (IIASA) citizen science platforms (LACO-Wiki and Geo-Wiki), as well as from individual project contributions. Data of 2016 onwards were collected, harmonised, and annotated. The data sets spatial, temporal, and thematic quality were assessed applying rules developed in this research. Currently, the repository holds around 75 million harmonised observations with standardized metadata of which a large share is available to the public. The repository, funded by ESA through the WorldCereal project, can be used for either the calibration of image classification deep learning algorithms or the validation of Earth Observation generated products, such as global cropland extent and maize and wheat maps. We recommend continuing and institutionalizing this reference data initiative e.g. through GEOGLAM, and encouraging the community to publish land cover and crop type data following the open science and open data principles.},

keywords = {},

pubstate = {published},

tppubtype = {article}

}

@article{nokey,

title = {Crop-Type Recognition in Street-level Images with Convolutional Neural Networks},

author = {Fernando Orduna-cabrera and Marcial Sandoval-Gastelum and Ian MCCALLUM and Linda SEE and Steffen Fritz and Santosh Karanam and Tobias Sturn and Valeria Javalera-Rincon and F. Fernando González-Navarro},

url = {https://www.preprints.org/manuscript/202307.0724/v1},

doi = {10.20944/preprints202307.0724.v1},

year  = {2023},

date = {2023-07-11},

urldate = {2023-07-11},

abstract = {The creation of crop-type maps from satellite data has proven challenging, often impeded by a lack of accurate in-situ data. This paper aims to demonstrate a method for crop-type (ie. Maize, Wheat and Other) recognition based on Convolutional Neural Networks using a bottom-up approach. We trained the model with a highly accurate dataset of crowdsourced labelled street-level imagery. Classification results achieved an AUC of 0.87 for wheat, 0.85 for maize and 0.73 for other. Given that wheat and maize are the two most common food crops globally, combined with an ever-increasing amount of available street-level imagery, this approach could help address the need for improved crop-type monitoring globally. Challenges remain in addressing the noisy aspect of street-level imagery (ie. buildings, hedgerows, automobiles, etc.), where a variety of different objects tend to restrict the view and confound the algorithms

},

keywords = {},

pubstate = {published},

tppubtype = {article}

}

@article{nokeyk,

title = {Monitoring direct drivers of small-scale tropical forest disturbance in near real-time with Sentinel-1 and -2 data},

author = {Bart Slagter and Johannes Reiche and Diego Marcos and Adugna Mullissa and Etse Lossou and Marielos Peña-Claros and Martin Herold},

url = {https://www.sciencedirect.com/science/article/pii/S0034425723002067?via%3Dihub},

doi = {10.1016/j.rse.2023.113655},

year  = {2023},

date = {2023-06-21},

volume = {295},

abstract = {Advancements in satellite-based forest monitoring increasingly enable the near real-time detection of small-scale tropical forest disturbances. However, there is an urgent need to enhance such monitoring methods with automated direct driver attributions to detected disturbances. This would provide important additional information to make forest disturbance alerts more actionable and useful for uptake by different stakeholders. In this study, we demonstrate spatially explicit and near real-time methods to monitor direct drivers of small-scale tropical forest disturbance across a range of tropical forest conditions in Suriname, the Republic of the Congo and the Democratic Republic of the Congo. We trained a convolutional neural network with Sentinel-1 and Sentinel-2 data to continuously classify newly detected RAdar for Detecting Deforestation (RADD) alerts as smallholder agriculture, road development, selective logging, mining or other. Different monitoring scenarios were evaluated based on varying sensor combinations, post-disturbance time periods and confidence levels. In general, the use of Sentinel-2 data was found to be most accurate for driver classifications, especially with data composited over a period of 4 to 6 months after the disturbance detection. Sentinel-1 data showed to be valuable for more rapid classifications of specific drivers, especially in areas with persistent cloud cover. Throughout all monitoring scenarios, smallholder agriculture was classified most accurately, while road development, selective logging and mining were more challenging to distinguish. An accuracy assessment throughout the full extent of our study regions revealed a Macro-F1 score of 0.861 and an Overall Accuracy of 0.897 for the best performing model, based on the use of 6-month post-disturbance Sentinel-2 composites. Finally, we addressed three specific monitoring use cases that relate to rapid law enforcement against illegal activities, ecological impact assessments and timely carbon emission reporting, by optimizing the trade-off in classification timeliness and confidence to reach required accuracies. Our findings demonstrate the strong capacities of high spatiotemporal resolution satellite data for monitoring direct drivers of small-scale forest disturbance, considering different user interests. The produced forest disturbance driver maps can be accessed via},

keywords = {},

pubstate = {published},

tppubtype = {article}

}

@article{nokeyj,

title = {Impact of land tenure on deforestation control and forest restoration in Brazilian Amazonia},

author = {Gilberto Camara and Rolf Simoes and Heloisa M Ruivo and Pedro R Andrade and Aline C Soterroni and Fernando M Ramos and Rafael G Ramos and Marluce Scarabello and Claudio Almeida and Ieda Sanches and Luis Maurano and Alexandre Coutinho and Julio Esquerdo and João Antunes and Adriano Venturieri and Marcos Adami},

url = {https://iopscience.iop.org/article/10.1088/1748-9326/acd20a},

doi = {10.1088/1748-9326/acd20a},

year  = {2023},

date = {2023-05-12},

abstract = {This study examines how land tenure constrains Brazil's ability to meet its deforestation control and forest restoration goals in its Amazonia biome. Our findings are based on an updated assessment of land tenure and land use in the region. Between 2019 and 2021, 44% of deforestation in Amazonia occurred in private lands, while forest removal in settlements ranged from 31% to 27% of the total. Deforestation in undesignated public lands increased from 11% in 2008 to 18% in 2021. Deforestation is highly concentrated, with 1% of properties accounting for 82.5% of forest cuts in 2021. In Amazonia, there is considerable non-compliance with the legal reserve provisions set by Brazil's Forest Code. Legal reserve deficits in private lands sum up to 18.17 Mha (million hectares), compared with 12.49 Mha of legal reserve surpluses. Even if all forest surpluses are offered in the forest credits market set in the Forest Code, farmers still need to restore 5.67 Mha to comply with the law. Large-scale cattle ranchers have a legal reserve deficit of 10.35 Mha (34% of their area). Most crop farming occurs in medium and large properties (4.63 Mha) with a large proportion of legal reserve deficits (45%). Given the political power and financial resources of large ranchers and crop producers, Brazil faces major challenges in inducing these farmers to meet their legal obligations. Therefore, Brazil needs to combine robust command-and-control strategies with market-based policies to achieve its deforestation and forest restoration goals. The government should tailor forest protection and restoration policies to the needs of different landowners, considering their land use practices, technical capacity, and financial resources.},

keywords = {},

pubstate = {published},

tppubtype = {article}

}

@article{nokey,

title = {High-precision datasets from monitoring stations based on eddy covariance measurements: what six years of quality evaluation process of ICOS ecosystem stations have to tell},

author = {Simone Sabbatini and Giacomo Nicolini and Bert Gielen and Maarten Op de Beeck and Fana Michilsens and Arne Iserbyt and Denis Loustau and Sébastien Lafont and Benjamin Loubet and Eleonora Canfora and Diego Polidori and Alessio Ribeca and Dario Papale},

url = {https://meetingorganizer.copernicus.org/EGU23/EGU23-16343.html},

doi = {10.5194/egusphere-egu23-16343},

year  = {2023},

date = {2023-04-24},

abstract = {ICOS (Integrated Carbon Observation System) is a Research Infrastructure aiming at getting a deeper understanding of the European Carbon balance by means of a network of monitoring stations, based on eddy covariance (EC) technique, spread out all over the European Continent, and continuously expanding. The Ecosystem Thematic Centre (ETC) coordinates the activities of ecosystem stations to ensure high-precision datasets and standardisation. The so-called Labelling procedure is made of two steps, conceived to guide the candidate stations to get the official ICOS label: the Step 1 is focused on the sensors’ setup and is structured as a discussion between the ETC and the station teams, while the Step 2 concerns the practical build-up of the station and the data evaluation. For stations with the stricter standards (so-called Class 1 and Class 2), some quality tests on the data are included: one on the EC data quality, two on the representativeness of the measured EC fluxes and one on the representativeness of the ancillary plots.



Currently 58 of 86 candidate stations completed the labelling procedure, of which 30 Class 1 and 2. The more common fixes agreed in Step 1 are changes in sonic orientation and height or location, to better deal with fetch and canopy inhomogeneities. In Step 2, apart from increasing the signal resolution and fixing some metadata, a further correction of the location/height of the sensors led to solving the remaining problems. Overall, two thirds of the stations passed the three EC tests at the first try (all the wetlands, 74% of the forests, 33% of the crops), pointing at the efficiency of the Step 1 evaluations, while the remaining ten didn’t pass one or more of the two other EC tests, testifying that some issues are only discoverable from proper data analysis. About one third of the stations didn’t pass the ancillary representativeness test, all of them over forests: the most common solution was to add or move one or more plots.



The results support the common knowledge that more complex ecosystems - not uniform canopy geometries, fast growing vegetation - are more likely to be affected by some data quality issue. This constitutes a crucial warning to researchers and technicians in the direction of properly considering the station characteristics when planning its setup and sampling design, as well as continuously checking the data produced, to ensure the production of high-precision datasets.},

keywords = {},

pubstate = {published},

tppubtype = {article}

}

@article{nokeyd,

title = {StrucNet: a global network for automated vegetation structure monitoring},

author = {Kim Calders and Benjamin Brede and Glenn Newnham and Darius Culvenor and John Armston and Harm Bartholomeus and Anne Griebel and Jodie Hayward and Samuli Junttila and Alvaro Lau and Shaun Levick and Rosalinda Morrone and Niall Origo and Marion Pfeifer and Jan Verbesselt and Martin Herold},

url = {https://zslpublications.onlinelibrary.wiley.com/doi/10.1002/rse2.333#pane-pcw-figures},

doi = {10.1002/rse2.333},

year  = {2023},

date = {2023-04-14},

urldate = {2023-04-14},

abstract = {Climate change and increasing human activities are impacting ecosystems and their biodiversity. Quantitative measurements of essential biodiversity variables (EBV) and essential climate variables are used to monitor biodiversity and carbon dynamics and evaluate policy and management interventions. Ecosystem structure is at the core of EBVs and carbon stock estimation and can help to inform assessments of species and species diversity. Ecosystem structure is also used as an indirect indicator of habitat quality and expected species richness or species community composition. Spaceborne measurements can provide large-scale insight into monitoring the structural dynamics of ecosystems, but they generally lack consistent, robust, timely and detailed information regarding their full three-dimensional vegetation structure at local scales. Here we demonstrate the potential of high-frequency ground-based laser scanning to systematically monitor structural changes in vegetation. We present a proof-of-concept high-temporal ecosystem structure time series of 5 years in a temperate forest using terrestrial laser scanning (TLS). We also present data from automated high-temporal laser scanning that can allow upscaling of vegetation structure scanning, overcoming the limitations of a typically opportunistic TLS measurement approach. Automated monitoring will be a critical component to build a network of field monitoring sites that can provide the required calibration data for satellite missions to effectively monitor the structural dynamics of vegetation over large areas. Within this perspective, we reflect on how this network could be designed and discuss implementation pathways.



},

keywords = {},

pubstate = {published},

tppubtype = {article}

}

@article{nokeyl,

title = {Past decade above-ground biomass change comparisons from four multi-temporal global maps},

author = {Arnan Araza and Martin Herold and Sytze de Bruin and Philippe Ciais and David A. Gibbs and Nancy Harris and Maurizio Santoro and Jean-Pierre Wigneron and Hui Yang and Natalia Málaga and Karimon Nesha and Pedro Rodriguez-Veiga and Olga Brovkina and Hugh C.A. Brown and Milen Chanev and Zlatomir Dimitrov and Lachezar Filchev and Jonas Fridman and Mariano García and Alexander Gikov and Leen Govaere and Petar Dimitrov and Fardin Moradi and Adriane Esquivel Muelbert and Jan Novotný and Thomas A.M. Pugh and Mart-Jan Schelhaas and Dmitry Schepaschenko and Krzysztof Stereńczak and Lars Hein},

url = {https://www.sciencedirect.com/science/article/pii/S1569843223000961?via%3Dihub},

doi = {10.1016/j.jag.2023.103274},

year  = {2023},

date = {2023-04-04},

volume = {118},

abstract = {Above-ground biomass (AGB) is considered an essential climate variable that underpins our knowledge and information about the role of forests in mitigating climate change. The availability of satellite-based AGB and AGB change (

AGB) products has increased in recent years. Here we assessed the past decade net 

AGB derived from four recent global multi-date AGB maps: ESA-CCI maps, WRI-Flux model, JPL time series, and SMOS-LVOD time series. Our assessments explore and use different reference data sources with biomass re-measurements within the past decade. The reference data comprise National Forest Inventory (NFI) plot data, local 

AGB maps from airborne LiDAR, and selected Forest Resource Assessment country data from countries with well-developed monitoring capacities. Map to reference data comparisons were performed at levels ranging from 100 m to 25 km spatial scale. The comparisons revealed that LiDAR data compared most reasonably with the maps, while the comparisons using NFI only showed some agreements at aggregation levels 

10 km. Regardless of the aggregation level, AGB losses and gains according to the map comparisons were consistently smaller than the reference data. Map-map comparisons at 25 km highlighted that the maps consistently captured AGB losses in known deforestation hotspots. The comparisons also identified several carbon sink regions consistently detected by all maps. However, disagreement between maps is still large in key forest regions such as the Amazon basin. The overall 

AGB map cross-correlation between maps varied in the range 0.11–0.29 (r). Reported 

AGB magnitudes were largest in the high-resolution datasets including the CCI map differencing (stock change) and Flux model (gain-loss) methods, while they were smallest according to the coarser-resolution LVOD and JPL time series products, especially for AGB gains. Our results suggest that 

AGB assessed from current maps can be biased and any use of the estimates should take that into account. Currently, 

AGB reference data are sparse especially in the tropics but that deficit can be alleviated by upcoming LiDAR data networks in the context of Supersites and GEO-Trees.},

keywords = {},

pubstate = {published},

tppubtype = {article}

}

@article{nokey,

title = {Forest restoration challenges in Brazilian Amazonia},

author = {Gilberto Camara and Rolf Simoes and Heloisa Ruivo and Pedro Andrade and Aline Soterroni and Fernando Ramos and Rafael Ramos and Marluce Scarabello and Claudio Almeida and Ieda Sanches and Luis Maurano and Alexandre Coutinho and Julio Esquerdo and Joao Antunes and Adriano Venturieri and Marcos Adami },

url = {https://www.preprints.org/manuscript/202303.0365/v1},

doi = {10.20944/preprints202303.0365.v1 },

year  = {2023},

date = {2023-03-21},

urldate = {2023-03-21},

abstract = {In its Nationally Determined Contribution (NDC) to the United Nations Framework Convention on Climate Change, Brazil committed to reducing greenhouse gas emissions and restoring its forests. This study examines the challenges of fulfilling these commitments in Brazilian Amazonia. We carry out a detailed assessment of the current status of land tenure in the region and its relation to deforestation. After dealing with conflicts and overlaps between data from various sources, we produce a new map of public and private land tenure in Amazonia. Combining this map with Brazil's official data on deforestation, we find out how much natural vegetation has been preserved in each public or private area. The result is used to estimate how much deforestation is illegal. We also establish how much deforestation is associated with each land tenure type. Our results show that most deforestation inside rural properties is done by a few landowners, a finding that has important consequences for law enforcement. We then assess the challenges for reforestation in detail. To do so, we consider how much forest needs to be rehabilitated according to Brazil's Forest Code. Our analysis provides a comprehensive appraisal of the potential opportunity costs for forest restoration in the biome, considering farm size and land use. This analysis provides insights into targeted land use policies that can meet Brazil’s forest restoration goals.

},

keywords = {},

pubstate = {published},

tppubtype = {article}

}

@article{nokey,

title = {G-reqs, a New Model Proposal for Capturing and Managing In Situ Data Requirements: First Results in the Context of the Group on Earth Observations},

author = {Joan Maso and Alba Brobia and Marie-Francoise Voidrot and Alaitz Zabala and Ivette Serral},

url = {https://www.mdpi.com/2072-4292/15/6/1589},

doi = {10.3390/rs15061589},

issn = {2072-4292},

year  = {2023},

date = {2023-03-15},

urldate = {2023-03-15},

journal = {Remote Sensing},

volume = {15},

number = {1589},

issue = {6},

abstract = {In the field of Earth observation, the importance of in situ data was recognized by the Group on Earth Observations (GEO) in the Canberra Declaration in 2019. The GEO community focuses on three global priority engagement areas: the United Nations 2030 Agenda for Sustainable Development, the Paris Agreement, and the Sendai Framework for Disaster Risk Reduction. While efforts have been made by GEO to open and disseminate in situ data, GEO did not have a general way to capture in situ data user requirements and drive the data provider efforts to meet the goals of its three global priorities. We present a requirements data model that first formalizes the collection of user requirements motivated by user-driven needs. Then, the user requirements can be grouped by essential variable and an analysis can derive product requirements and parameters for new or existing products. The work was inspired by thematic initiatives, such as OSCAR, from WMO, OSAAP (formerly COURL and NOSA) from NOAA, and the Copernicus In Situ Component Information System. The presented solution focuses on requirements for all applications of Earth observation in situ data. We present initial developments and testing of the data model and discuss the steps that GEO should take to implement a requirements database that is connected to actual data in the GEOSS platform and propose some recommendations on how to articulate it.},

keywords = {},

pubstate = {published},

tppubtype = {article}

}

@article{nokeyg,

title = {A widely-used eddy covariance gap-filling method creates systematic bias in carbon balance estimates},

author = {Henriikka Vekuri and Juha-Pekka Tuovinen and Liisa Kulmala and Dario Papale and Pasi Kolari and Mika Aurela and Tuomas Laurila and Jari Liski and Annalea Lohila },

url = {https://www.nature.com/articles/s41598-023-28827-2#additional-information},

doi = {10.1038/s41598-023-28827-2},

year  = {2023},

date = {2023-01-31},

abstract = {Climate change mitigation requires, besides reductions in greenhouse gas emissions, actions to increase carbon sinks in terrestrial ecosystems. A key measurement method for quantifying such sinks and calibrating models is the eddy covariance technique, but it requires imputation, or gap-filling, of missing data for determination of annual carbon balances of ecosystems. Previous comparisons of gap-filling methods have concluded that commonly used methods, such as marginal distribution sampling (MDS), do not have a significant impact on the carbon balance estimate. By analyzing an extensive, global data set, we show that MDS causes significant carbon balance errors for northern (latitude >60∘

) sites. MDS systematically overestimates the carbon dioxide (CO2

) emissions of carbon sources and underestimates the CO2

 sequestration of carbon sinks. We also reveal reasons for these biases and show how a machine learning method called extreme gradient boosting or a modified implementation of MDS can be used to substantially reduce the northern site bias.},

keywords = {},

pubstate = {published},

tppubtype = {article}

}

@article{@article{NESHA2022157788,

title = {Exploring characteristics of national forest inventories for integration with global space-based forest biomass data},

author = {Karimon Nesha and Martin Herold and Veronique {De Sy} and Sytze {de Bruin} and Arnan Araza and Natalia Málaga and Javier G.P. Gamarra and Kristell Hergoualc'h and Anssi Pekkarinen and Carla Ramirez and David Morales-Hidalgo and Rebecca Tavani},

url = {https://www.sciencedirect.com/science/article/pii/S0048969722048872?via%3Dihub},

doi = {https://doi.org/10.1016/j.scitotenv.2022.157788},

year  = {2022},

date = {2022-12-01},

journal = {Science of The Total Environment},

volume = {850},

issue = {157788},

abstract = {National forest inventories (NFIs) are a reliable source for national forest measurements. However, they are usually not developed for linking with remotely sensed (RS) biomass information. There are increasing needs and opportunities to facilitate this link towards better global and national biomass estimation. Thus, it is important to study and understand NFI characteristics relating to their integration with space-based products; in particular for the tropics where NFIs are quite recent, less frequent, and partially incomplete in several countries. Here, we (1) assessed NFIs in terms of their availability, temporal distribution, and extent in 236 countries from FAO's Global Forest Resources Assessment (FRA) 2020; (2) compared national forest biomass estimates in 2018 from FRA and global space-based Climate Change Initiative (CCI) product in 182 countries considering NFI availability and temporality; and (3) analyzed the latest NFI design characteristics in 46 tropical countries relating to their integration with space-based biomass datasets. We observed significant NFI availability globally and multiple NFIs were mostly found in temperate and boreal countries while most of the single NFI countries (94 %) were in the tropics. The latest NFIs were more recent in the tropics and many countries (35) implemented NFIs from 2016 onwards. The increasing availability and update of NFIs create new opportunities for integration with space-based data at the national level. This is supported by the agreement we found between country biomass estimates for 2018 from FRA and CCI product, with a significantly higher correlation in countries with recent NFIs. We observed that NFI designs varied greatly in tropical countries. For example, the size of the plots ranged from 0.01 to 1 ha and more than three-quarters of the countries had smaller plots of ≤0.25 ha. The existing NFI designs could pose specific challenges for statistical integration with RS data in the tropics. Future NFI and space-based efforts should aim towards a more integrated approach taking advantage of both data streams to improve national estimates and help future data harmonization efforts. Regular NFI efforts can be expanded with the inclusion of some super-site plots to enhance data integration with currently available space-based applications. Issues related to cost implications versus improvements in the accuracy, timeliness, and sustainability of national forest biomass estimation should be further explored.},

keywords = {},

pubstate = {published},

tppubtype = {article}

}

@article{nokeym,

title = {Challenges to Achieving the Commitments of Brazil’s NDC in the Amazon Biome},

author = {Gilberto Câmara},

url = {https://cebri.org/revista/en/artigo/63/challenges-to-achieving-the-commitments-of-brazils-ndc-in-the-amazon-biome},

year  = {2022},

date = {2022-11-28},

abstract = {The Brazilian Nationally Determined Contribution (NDC) established ambitious greenhouse gas emissions reduction targets. Using the Brazilian Forestry Code as reference, we describe the challenges Brazil will face to achieve its deforestation and forest restoration NDC commitments in the Amazon biome. This paper proposes a new perspective for public policy focusing on the extent of illegal deforestation and on current land use in the region.



},

keywords = {},

pubstate = {published},

tppubtype = {article}

}

@article{nokeyh,

title = {Precision of subnational forest AGB estimates within the Peruvian Amazonia using a global biomass map},

author = {Natalia Málaga and Sytze de Bruin and Ronald E. McRoberts and Alexs Arana Olivos and Ricardo de la Cruz Paiva and Patricia Durán Montesinos and Daniela Requena Suarez and Martin Herold},

url = {https://www.sciencedirect.com/science/article/pii/S1569843222002904?via%3Dihub},

doi = {10.1016/j.jag.2022.103102},

year  = {2022},

date = {2022-11-18},

volume = {115},

abstract = {National forest inventories (NFI) provide essential forest-related biomass and carbon information for country greenhouse gas (GHG) accounting systems. Several tropical countries struggle to execute their NFIs while the extent to which space-based global information on aboveground biomass (AGB) can support national GHG accounting is under investigation. We assess whether the use of a global AGB map as auxiliary information produces a gain in precision of subnational AGB estimates for the Peruvian Amazonia. We used model-assisted estimators with data from the country’s NFI and explored hybrid inferential techniques to account for the sources of uncertainty associated with the integration of remote sensing-based products and NFI plot data.



Our results show that the selected global biomass map tends to overestimate AGB values across the Peruvian Amazonia. For most strata, directly using the map in its published form did not reduce the precision of AGB estimates. However, after calibrating the map using the NFI data, the precision of our map-assisted AGB estimates increased by up to 50% at stratum level and 20% at Amazonia level. We further demonstrate how different sources of uncertainties can be incorporated in the map-NFI integrated estimates. With the hybrid inferential analysis, we found that the small spatial support of the NFI plots compared to the remote sensing-based sample units of aggregated pixels (within block variability) contributed the most to the total uncertainty associated with the AGB estimates from our map-NFI integration. Uncertainties caused by measurement variability and allometric model prediction uncertainty were the second largest contributors. When these uncertainties were incorporated, the increase in precision of our calibrated map-assisted AGB estimates was negligible, probably hindered by the great contribution of the within block variability to our map-plot assessment. We developed a reproducible method that countries can build upon and further improve while the global biomass products continue to evolve and better characterize the AGB distribution under large biomass conditions. We encourage further cross-country case studies that reflect a wider range of AGB distributions, especially within humid tropical forests, to further assess the contribution of global biomass maps to (sub)national AGB estimates and finally GHG monitoring and reporting.},

keywords = {},

pubstate = {published},

tppubtype = {article}

}

@article{nokey,

title = {A Downsampling Method Addressing the Modifiable Areal Unit Problem in Remote Sensing},

author = {Andrei Mîrț and Johannes Reiche and Jan Verbesselt and Martin Herold},

url = {https://www.mdpi.com/2072-4292/14/21/5538},

doi = {https://doi.org/10.3390/rs14215538},

year  = {2022},

date = {2022-11-03},

journal = {Remote Sensing},

volume = {14},

number = {5538},

issue = {22},

abstract = {Handling multiple scales efficiently is one avenue for processing big remote sensing imagery data. Unfortunately, imagery is also affected by the infamous modifiable areal unit problem, which creates unpredictable errors at different scales. We developed a downsampling method that attempts to keep the data distribution in a downsampled image constant, reducing the modifiable areal unit problem. We tested our method against classic downsampling methods (mean, central pixel selection, random) under a range of typical remote sensing scenarios. Under our experimental conditions, our downsampling method consistently outperformed the classical downsampling methods within a 95% confidence level. The downsampling method can be used in most typical situations where downsampling is needed, but it is likely to shine when used as a pyramid building policy in geocomputing platforms, such as Google Earth Engine.},

keywords = {},

pubstate = {published},

tppubtype = {article}

}

@article{nokeye,

title = {Toward a forest biomass reference measurement system for remote sensing applications},

author = {Nicolas Labrière and Stuart J. Davies and Mathias I. Disney and Laura I. Duncanson and Martin Herold and Simon L. Lewis and Oliver L. Phillips and Shaun Quegan and Sassan S. Saatchi and Dmitry G. Schepaschenko and Klaus Scipal and Plinio Sist and Jérôme Chave},

url = {https://onlinelibrary.wiley.com/doi/10.1111/gcb.16497},

doi = {10.1111/gcb.16497},

year  = {2022},

date = {2022-10-21},

volume = {29},

issue = {3},

pages = {827-840},

abstract = {Forests contribute to climate change mitigation through carbon storage and uptake, but the extent to which this carbon pool varies in space and time is still poorly known. Several Earth Observation missions have been specifically designed to address this issue, for example, NASA's GEDI, NASA-ISRO's NISAR and ESA's BIOMASS. Yet, all these missions' products require independent and consistent validation. A permanent, global, in situ, site-based forest biomass reference measurement system relying on ground data of the highest possible quality is therefore needed. Here, we have assembled a list of almost 200 high-quality sites through an in-depth review of the literature and expert knowledge. In this study, we explore how representative these sites are in terms of their coverage of environmental conditions, geographical space and biomass-related forest structure, compared to those experienced by forests worldwide. This work also aims at identifying which sites are the most representative, and where to invest to improve the representativeness of the proposed system. We show that the environmental coverage of the system does not seem to improve after at least the 175 most representative sites are included, but geographical and structural coverages continue to improve as more sites are added. We highlight the areas of poor environmental, geographical, or structural coverage, including, but not limited to, Canada, the western half of the USA, Mexico, Patagonia, Angola, Zambia, eastern Russia, and tropical and subtropical highlands (e.g. in Colombia, the Himalayas, Borneo, Papua). For the proposed system to succeed, we stress that (1) data must be collected and processed applying the same standards across all countries and continents; (2) system establishment and management must be inclusive and equitable, with careful consideration of working conditions; and (3) training and site partner involvement in downstream activities should be mandatory.},

keywords = {},

pubstate = {published},

tppubtype = {article}

}

@article{nokey,

title = {Ways forward for Machine Learning to make useful global environmental datasets from legacy observations and measurements},

author = {Tomislav Hengl },

doi = {https://doi.org/10.1038/s41467-022-32693-3},

year  = {2022},

date = {2022-09-07},

urldate = {2022-09-07},

journal = {Nature Communications},

volume = {13},

number = {5178},

abstract = {Advances in geospatial and Machine Learning techniques for large datasets of georeferenced observations have made it possible to produce model-based global maps of ecological and environmental variables. However, the implementation of existing scientific methods (especially Machine Learning models) to produce accurate global maps is often complex. Tomislav Hengl (co-founder of OpenGeoHub foundation), Johan van den Hoogen (researcher at ETH Zürich), and Devin Routh (Science IT Consultant at the University of Zürich) shared with Nature Communications their perspectives for creators and users of these maps, focusing on the key challenges in producing global environmental geospatial datasets to achieve significant impacts.},

keywords = {},

pubstate = {published},

tppubtype = {article}

}

@article{@article{doi:10.1080/15481603.2022.2115619,

title = {Using high-resolution imagery and deep learning to classify land-use following deforestation: a case study in Ethiopia},

author = {Robert N. Masolele and Veronique De Sy and Diego Marcos and Jan Verbesselt and Fabian Gieseke and Kalkidan Ayele Mulatu and Yitebitu Moges and Heiru Sebrala and Christopher Martius and Martin Herold},

url = {https://doi.org/10.1080/15481603.2022.2115619

},

doi = {https://doi.org/10.1080/15481603.2022.2115619},

year  = {2022},

date = {2022-09-07},

urldate = {2022-09-07},

journal = {GIScience & Remote Sensing},

volume = {59},

number = {1},

pages = {1446-1472},

abstract = {National-scale assessments of post-deforestation land-use are crucial for decreasing deforestation and forest degradation-related emissions. In this research, we assess the potential of different satellite data modalities (single-date, multi-date, multi-resolution, and an ensemble of multi-sensor images) for classifying land-use following deforestation in Ethiopia using the U-Net deep neural network architecture enhanced with attention. We performed the analysis on satellite image data retrieved across Ethiopia from freely available Landsat-8, Sentinel-2 and Planet-NICFI satellite data. The experiments aimed at an analysis of (a) single-date images from individual sensors to account for the differences in spatial resolution between image sensors in detecting land-uses, (b) ensembles of multiple images from different sensors (Planet-NICFI/Sentinel-2/Landsat-8) with different spatial resolutions, (c) the use of multi-date data to account for the contribution of temporal information in detecting land-uses, and, finally, (d) the identification of regional differences in terms of land-use following deforestation in Ethiopia. We hypothesize that choosing the right satellite imagery (sensor) type is crucial for the task. Based on a comprehensive visually interpreted reference dataset of 11 types of post-deforestation land-uses, we find that either detailed spatial patterns (single-date Planet-NICFI) or detailed temporal patterns (multi-date Sentinel-2, Landsat-8) are required for identifying land-use following deforestation, while medium-resolution single-date imagery is not sufficient to achieve high classification accuracy. We also find that adding soft-attention to the standard U-Net improved the classification accuracy, especially for small-scale land-uses. The models and products presented in this work can be used as a powerful data resource for governmental and forest monitoring agencies to design and monitor deforestation mitigation measures and data-driven land-use policy.},

keywords = {},

pubstate = {published},

tppubtype = {article}

}