Dataset Information

Estimating herbaceous aboveground biomass in Sahelian rangelands using Structure from Motion data collected on the ground and by UAV.

ABSTRACT: Herbaceous aboveground biomass (HAB) is a key indicator of grassland vegetation and indirect estimation tools, such as remote sensing imagery, increase the potential for covering larger areas in a timely and cost-efficient way. Structure from Motion (SfM) is an image analysis process that can create a variety of 3D spatial models as well as 2D orthomosaics from a set of images. Computed from Unmanned Aerial Vehicle (UAV) and ground camera measurements, the SfM potential to estimate the herbaceous aboveground biomass in Sahelian rangelands was tested in this study. Both UAV and ground camera recordings were used at three different scales: temporal, landscape, and national (across Senegal). All images were processed using PIX4D software (photogrammetry software) and were used to extract vegetation indices and heights. A random forest algorithm was used to estimate the HAB and the average estimation errors were around 150 g m^-² for fresh mass (20% relative error) and 60 g m^-² for dry mass (around 25% error). A comparison between different datasets revealed that the estimates based on camera data were slightly more accurate than those from UAV data. It was also found that combining datasets across scales for the same type of tool (UAV or camera) could be a useful option for monitoring HAB in Sahelian rangelands or in other grassy ecosystems.

SUBMITTER: Taugourdeau S

PROVIDER: S-EPMC9057245 | biostudies-literature | 2022 May

REPOSITORIES: biostudies-literature

ACCESS DATA

Publications

Estimating herbaceous aboveground biomass in Sahelian rangelands using Structure from Motion data collected on the ground and by UAV.

Taugourdeau Simon S Diedhiou Antoine A Fassinou Cofélas C Bossoukpe Marina M Diatta Ousmane O N'Goran Ange A Auderbert Alain A Ndiaye Ousmane O Diouf Abdoul Aziz AA Tagesson Torbern T Fensholt Rasmus R Faye Emile E

Ecology and evolution 20220501 5

Herbaceous aboveground biomass (HAB) is a key indicator of grassland vegetation and indirect estimation tools, such as remote sensing imagery, increase the potential for covering larger areas in a timely and cost-efficient way. Structure from Motion (SfM) is an image analysis process that can create a variety of 3D spatial models as well as 2D orthomosaics from a set of images. Computed from Unmanned Aerial Vehicle (UAV) and ground camera measurements, the SfM potential to estimate the herbaceou ...[more]

PMID: 35509616

Similar Datasets

Project description:BackgroundAbove-ground biomass (AGB) is a basic agronomic parameter for field investigation and is frequently used to indicate crop growth status, the effects of agricultural management practices, and the ability to sequester carbon above and below ground. The conventional way to obtain AGB is to use destructive sampling methods that require manual harvesting of crops, weighing, and recording, which makes large-area, long-term measurements challenging and time consuming. However, with the diversity of platforms and sensors and the improvements in spatial and spectral resolution, remote sensing is now regarded as the best technical means for monitoring and estimating AGB over large areas.ResultsIn this study, we used structural and spectral information provided by remote sensing from an unmanned aerial vehicle (UAV) in combination with machine learning to estimate maize biomass. Of the 14 predictor variables, six were selected to create a model by using a recursive feature elimination algorithm. Four machine-learning regression algorithms (multiple linear regression, support vector machine, artificial neural network, and random forest) were evaluated and compared to create a suitable model, following which we tested whether the two sampling methods influence the training model. To estimate the AGB of maize, we propose an improved method for extracting plant height from UAV images and a volumetric indicator (i.e., BIOVP). The results show that (1) the random forest model gave the most balanced results, with low error and a high ratio of the explained variance for both the training set and the test set. (2) BIOVP can retain the largest strength effect on the AGB estimate in four different machine learning models by using importance analysis of predictors. (3) Comparing the plant heights calculated by the three methods with manual ground-based measurements shows that the proposed method increased the ratio of the explained variance and reduced errors.ConclusionsThese results lead us to conclude that the combination of machine learning with UAV remote sensing is a promising alternative for estimating AGB. This work suggests that structural and spectral information can be considered simultaneously rather than separately when estimating biophysical crop parameters.

Project description:Worldwide savannas and arid grasslands are mainly used for livestock grazing, providing livelihood to over a billion people. While normally dominated by perennial C4 grasses, these rangelands are increasingly affected by the massive spread of native, mainly woody legumes. The consequences are often a repression of grass cover and productivity, leading to a reduced carrying capacity. While such encroachment by woody plants has been extensively researched, studies on similar processes involving herbaceous species are rare. We studied the impact of a sustained and massive spread of the native herbaceous legume Crotalaria podocarpa in Namibia's escarpment region on the locally dominant fodder grasses Stipagrostis ciliata and Stipagrostis uniplumis. We measured tussock densities, biomass production of individual tussocks and tussock dormancy state of Stipagrostis on ten 10 m x 10 m plots affected and ten similarly-sized plots unaffected by C. podocarpa over eight consecutive years and under different seasonal rainfalls and estimated the potential relative productivity of the land. We found the percentage of active Stipagrostis tussocks and the biomass production of individual tussocks to increase asymptotically with higher seasonal rainfall reaching a maximum around 300 mm while the land's relative productivity under average local rainfall conditions reached only 40% of its potential. Crotalaria podocarpa encroachment had no effect on the proportion of productive grass tussocks, but reduced he productivity of individual Stipagrostis tussocks by a third. This effect of C. podocarpa on grass productivity was immediate and direct and was not compensated for by above-average rainfall. Besides this immediate effect, over time, the density of grass tussocks declined by more than 50% in areas encroached by C. podocarpa further and lastingly reducing the lands carrying capacity. The effects of C. podocarpa on grass productivity hereby resemble those of woody encroachers. Therefore, against the background of global change, the spread of herbaceous legumes and the underlying patterns needs to be further investigated to develop adequate counter measures for a sustainable land use.

Project description:As countries advance in greenhouse gas (GHG) accounting for climate change mitigation, consistent estimates of aboveground net biomass change (∆AGB) are needed. Countries with limited forest monitoring capabilities in the tropics and subtropics rely on IPCC 2006 default ∆AGB rates, which are values per ecological zone, per continent. Similarly, research into forest biomass change at a large scale also makes use of these rates. IPCC 2006 default rates come from a handful of studies, provide no uncertainty indications and do not distinguish between older secondary forests and old-growth forests. As part of the 2019 Refinement to the 2006 IPCC Guidelines for National Greenhouse Gas Inventories, we incorporate ∆AGB data available from 2006 onwards, comprising 176 chronosequences in secondary forests and 536 permanent plots in old-growth and managed/logged forests located in 42 countries in Africa, North and South America and Asia. We generated ∆AGB rate estimates for younger secondary forests (≤20 years), older secondary forests (>20 years and up to 100 years) and old-growth forests, and accounted for uncertainties in our estimates. In tropical rainforests, for which data availability was the highest, our ∆AGB rate estimates ranged from 3.4 (Asia) to 7.6 (Africa) Mg ha-1 year-1 in younger secondary forests, from 2.3 (North and South America) to 3.5 (Africa) Mg ha-1 year-1 in older secondary forests, and 0.7 (Asia) to 1.3 (Africa) Mg ha-1 year-1 in old-growth forests. We provide a rigorous and traceable refinement of the IPCC 2006 default rates in tropical and subtropical ecological zones, and identify which areas require more research on ∆AGB. In this respect, this study should be considered as an important step towards quantifying the role of tropical and subtropical forests as carbon sinks with higher accuracy; our new rates can be used for large-scale GHG accounting by governmental bodies, nongovernmental organizations and in scientific research.

Dataset Information

Estimating herbaceous aboveground biomass in Sahelian rangelands using Structure from Motion data collected on the ground and by UAV.

Publications

Estimating herbaceous aboveground biomass in Sahelian rangelands using Structure from Motion data collected on the ground and by UAV.

Similar Datasets

OmicsDI is part of the ELIXIR infrastructure

Tweets