1,720,992 research outputs found
Evaluation of multi-temporal and multi-sensor atmospheric correction strategies for land cover accounting and monitoring in Ireland
Full text linkAccurate atmospheric correction is an important pre-processing step for studies of multitemporal
landcover mapping using optical satellite data. Model-based surface reflectance
predictions (e.g. 6S - Second Simulation of Satellite Signal in the Solar Spectrum) are highly
dependent on the adjustment of aerosol optical thickness (AOT) data. For regions with no or
insufficient spatial and temporal coverage of meteorological ground measurements, MODIS
derived AOT data are a valuable alternative, especially with regard to the dynamics of
atmospheric conditions. In this study, atmospheric correction strategies were assessed based on
the change in standard deviation (σ) compared to the raw data and also by machine learning
landcover classification accuracies. For three Landsat 8 OLI (acquired in 2013) and two
RapidEye (acquired in 2010 and 2014) scenes, seven different correction strategies were tested
over an agricultural area in south-east Ireland. Visibility calculated from daily spatial averaged
TERRA-MODIS estimates (1° × 1° Aerosol Product) served as input for the atmospheric
correction. In almost all cases the standard deviation of the raw data is reduced after
incorporation of terrain correction, compared to the atmospheric corrected data. ATCOR®-IDL
based correction decreases the standard deviation almost consistently (ranging from -0.3 to -
26.7). The 6S implementation in GRASS GIS showed a tendency of increasing the variation in
the data, especially for the RapidEye data. No major differences in overall accuracies and
Kappa values were observed between the three machine learning classification approaches. The
results indicate that the ATCOR®-IDL based correction and MODIS parametrisation methods
are able to decrease the standard deviation and are therefore an appropriate approach to
approximate the top-of-canopy reflectance
Upland vegetation mapping using Random Forests with optical and radar satellite data
Full text linkUplands represent unique landscapes that provide a range of vital benefits to society, but are under increasing pressure from the management needs of a diverse number of stakeholders (e.g. farmers, conservationists, foresters, government agencies and recreational users). Mapping the spatial distribution of upland vegetation could benefit management and conservation programmes and allow for the impacts of environmental change (natural and anthropogenic) in these areas to be reliably estimated. The aim of this study was to evaluate the use of medium spatial resolution optical and radar satellite data, together with ancillary soil and topographic data, for identifying and mapping upland vegetation using the Random Forests (RF) algorithm. Intensive field survey data collected at three study sites in Ireland as part of the National Parks and Wildlife Service (NPWS) funded survey of upland habitats was used in the calibration and validation of different RF models. Eight different datasets were analysed for each site to compare the change in classification accuracy depending on the input variables. The overall accuracy values varied from 59.8% to 94.3% across the three study locations and the inclusion of ancillary datasets containing information on the soil and elevation further improved the classification accuracies (between 5 and 27%, depending on the input classification dataset). The classification results were consistent across the three different study areas, confirming the applicability of the approach under different environmental contexts
A Framework for Satellite Image Classification in the Context of Crisis Mapping Using Markov Random Fields
Full text linkIn this contribution a framework for classification of high resolution optical satellite images in the context of crisis mapping is proposed and evaluated. Multiscale image information (data model) as well as hierarchical and spatial context information (prior model) is incorporated into the classification process using a hybrid Markov model which combines a hierarchical directed as well as a planar lattice-based Markov Random Field (MRF). The modelling of arbitrary semantic classes in different scales enables the definition of a hierarchical semantic network representing the dependencies and relations between the classes in adjacent scales. Classification is carried out using non-iterative hierarchical maximum a-posteriori (MAP) or mode of posterior marginal (MPM) inference as well as a subsequent optimization step using a planar MRF. Additionally, a modified MAP estimation which is able to outperform the original estimators under certain conditions is proposed. The impact of incorporation of image data from multiple scales is evaluated in this contribution. Furthermore, the dependency between the quantity of the training data and the classification accuracy is analyzed
The monitoring and modelling of the impacts of storms under sea-level rise on a breached coastal dune-barrier system
No full textLittle is known about the impacts of storms on breached barriers, and virtually nothing is known about the impacts of storms under a rising sea-level on these systems. This PhD research aims to help fill this gap. In 2008, barrier breaching at Rossbehy, Co. Kerry resulted in the establishment of a new tidal inlet. Semidiurnal tidal exchange through the new channel has been on going since this event. Rossbehy provides an excellent opportunity to study the influence of storms on barrier evolution post-breaching. A two-year monitoring campaign was undertaken to assess the morphological impacts of storms on Rossbehy and a neighbouring barrier, Inch. Multiple topographic surveys were conducted using terrestrial laser scanning (TLS) technology. The logistics of collecting, storing, processing, and analyzing this data were addressed in this research project. Major volume losses were recorded over the duration of the monitoring period at Rossbehy, but not at Inch. This difference was likely due to the orientation of the sites in relation to the main inlet channel. Meteorological data and numerically simulated nearshore wave data were used to identify and characterize storm events that occurred during the monitoring period. Strong negative statistically significant correlations were observed between rates of dune volume change and storm duration for events that occurred during the monitoring period. Additional statistical analyses revealed that event duration in combination with maximum significant wave height were the best predictors of dune volume change at Rossbehy. A novel experiment was set up to assess the impacts of storms under future sealevel rise (SLR) on Rossbehy using numerical modelling and TLS data. Numerical modelling was performed in MIKE21. TLS data was used to evaluate the effectiveness of the model in simulating dune volume changes near the breach. The results of the experiment indicate that under future SLR, storms will contribute to a net offshore movement of sediment in the near shore zone of Rossbehy. This will inevitably lead to shoreline retreat and could result in the possible drowning of the barrier if back barrier saltmarsh sediments cannot accumulate fast enough to keep up with rising sea-level. Based on the results of the monitoring campaign and modelling experiments, a conceptual model of the evolution of the system was developed – the S-SLR model. The model integrates the influence of storms under a rising sea-level into a previously developed conceptual model put forth by O’Shea (2015). The new model accounts for sediment deficits in the near shore zone caused by storms under a rising sea-level. This is the first assessment of the potential impacts of storms under sea-level rise on a breached barrier system in Ireland. It is envisaged that this study will serve as baseline from which to compare future process studies of similar systems
Urban green and blue space in Cork city and its importance to bird diversity
Full text linkUrban green and blue spaces are well known for providing ecosystem services in built up areas but they are also becoming increasingly important as these spaces are becoming unique ecosystems for bird diversity. The aims of this study was to investigate the importance of green and blue space in Cork City to bird diversity, through the generation of a satellite image that captures the landscape configuration of Cork City and green and blue space it contains, examine the relationships between bird diversity (richness and abundance) and landscape metrics generated using FRAGSATS and by conducting field surveys. The impact of spatial scales and how they affect species-landscape relationships was also investigated through regression analysis. Using Sentinel-2 satellite data and a maximum likelihood classification, a comprehensive landcover map of Cork City was produced with reliable accuracy. The map revealed that two thirds of the city is composed of green and blue space. The field surveys
recorded 62 species in the city. The statistical analysis gathered revealed that green space was the main driver in increasing species richness and abundance, while blue space produced mixed results. The edge effect phenomenon was also found to play a key role in increasing bird diversity. The regression models produced results that revealed a diversified and varied landscape was preferable to bird diversity as the scale was increased. The impact of scale also affected how important blue space is as a connective network within the city. Overall, this study has demonstrated that urban green and blue space is intrinsically linked to bird diversity in Cork City. 40% of the species that were recorded in the field surveys are listed as species of conservation concern in Ireland, with five of these species listed on the Red list. This finding has shown how urban spaces can provide habitats for vulnerable species, and provides precedence for implementing conservation initiatives within urban areas
Incorporating biotic interactions in phenology
No full textShifts in the timing of phenological events such as bird migration, leaf unfolding, flowering, and insect emergence, across many taxa and ecosystems are a result of climate change. Phenological shifts depend on different factors and species-specific sensitivity to changes in meteorological variables, therefore when phenological shifts occur within the trophic network we might expect phenological mismatches between interlinked species to occur as a result of climate change, with potential negative effects for biodiversity, ecosystems and the trophic network. However, the availability of data that show how species interactions are affected by climate change is scarce and unified criteria are still lacking on the methodologies studying phenology and biotic interactions. The presented extensive review on the topic allowed the identification of four broad categories of studies that have explored biotic interactions within phenology research and revealed that phenological studies of seasons other than spring are very scarce. This unbalance was also found within biotic interactions research, where mutualistic and obligate interactions, trophic interactions and networks were the main types receiving the most attention compared to other types (i.e., facilitation, competition). Researchers have commonly used co-existence among species as a proxy for biotic interactions, in many cases without any direct measurement of such interactions, while a lack of formal examination in most studies exploring phenological mismatches in response to climate change was also often identified. A conceptual framework was developed for the inclusion of phenology in the study of biotic interactions that categorises research into the conceptualisation and modelling of biotic interactions. Conceptualisation explores phenological data, types of interactions, and the spatiotemporal dimensions, which all determine the representation for biotic interactions within the modelling framework, and the type of models that are applicable. Emerging opportunities were also identified to investigate biotic interactions in phenology research, including spatially and temporally explicit species distribution models as proxies for phenological events and the combination of novel technologies (e.g., acoustic recorders, telemetry data) to quantify interactions. This conceptual framework was applied to a case of study in Ireland, investigating the relevance of different meteorological drivers (maximum and minimum temperature and total precipitation) in the phenology and co-existence of several species linked through the trophic network. Phenological trends towards an earlier phenology in Ireland were identified in terms of advanced date of arrival of migrant birds, first flight of butterflies and moths and green-up (start of the growing season) over the period 2008-2018. A novel analysis developed by van de Pol et al. (2016), the relative sliding time window analyses, was applied in order to identify which meteorological drivers had higher influence on the phenological events of study. Results showed high interannual variability in the time windows at species and group level. We identified common trends between butterflies and moths to show greater influence of temperature time windows when closer to first flight, while in vegetation the opposite pattern was found. Three new indices of phenological change across different trophic levels are presented, these indices allowed to identify potential phenological asynchronies between trophic levels in Ireland and to develop a network of potential interactions based on synchrony among interlinked species
National farm scale estimates of grass yield from satellite remote sensing
Full text linkGlobally, grasslands are an important source of food for livestock and provide additional ecosystem services such as greenhouse gas (GHG) mitigation through carbon sequestration, habitats for biodiversity, and recreational amenities. Grass is the cheapest source of fodder providing Irish farmers with an economic benefit against international competitors. Hence, to maintain profitability, farmers have to maximize the proportion of grazed grass in cow’s diet or save it as silage.
The overall objective of the current research project was to build a machine-learning model to estimate grass growth nationally using earth observation imagery from the Sentinel 2 satellite constellation and ancillary meteorological data, which are known to influence grass growth. Firstly, the impact of meteorological data and Growing Degree Days (GDD) was assessed for Teagasc Moorepark experimental farm (Fermoy, Co Cork, Ireland). GDD was modified to include Soil Moisture Deficit (SMD), which included the impact of summer drought conditions in 2018. Results demonstrated the importance of GDD for grass growth estimation using ordinary linear regression (OLS). The potential evapotranspiration (PE) 0.65 (r=0.65) and evaporation (r=0.65) were equally significant variables in 2017, while in 2018 the solar radiation had the highest correlation (r=0.43), followed by potential evapotranspiration and evaporation with r of 0.42. The standard and modified GDD were equally significant variables with r of 0.65 in 2017, but both had a reduced correlation in 2018 with modified GDD (0.38, p<0.01) performing slightly better than the standard GDD (0.26, p<0.01) calculation. These models only explained 53% (RMSE of 18.90 kg DM ha-1day-1) and 36% (RMSE of 27.02 kg DM ha-1day-1) of variability in grass growth for 2017 and 2018, respectively.
Considering the importance of meteorological data, an empirical grass model called the Brereton model, previously used for Irish grass growing conditions were tested. Since this model lacks a spatial element, we compared the Brereton model with the previously used machine-learning model ANFIS and Random Forest (RF) with the combination of satellite data and meteorological data for eight Teagasc farms. Overall, the machine-learning algorithms (R2= 0.32 to 0.73 and RMSE=14.65 to 24.76 kg DM ha-1day-1 for the test data) performed better than the Brereton model (range of R2=0.03 to 0.33 and RMSE=41.68 to 82.29 kg DM ha-1day-1). The RF model (with all the variables except rainfall) had the highest accuracy for predicting grass growth rate, with (R2= 0.55, RMSE = 14.65 kg DM ha-1day-1, MSE= 214.79 kg DM ha-1day-1 versus ANFIS with R2 = 0.47, RMSE = 15.95 kg DM ha-1day-1, MSE= 254.40 kg DM ha-1day-1).
When developing a national model, meteorological data were missing (except precipitation). A different approach was followed, whereby the grass growing season was subdivided (January-June Agmodel 1 and July–December Agmodel 2). Phenologically, the peak grass growth in Ireland typically occurs in May, with a slow decline in subsequent months. Spring is the most important season for grassland management, where growing conditions can impact the grass supply for the whole year. The national models were developed using Sentinel 2 band metrics, spectral indices (NDVI and NDRE), and rainfall for 179 farms. Data from 2017-2019 was divided into training and testing data (70:30 split), with 2020 data used for independent validation of the final trained model. Test accuracy was higher for Agmodel 1 (R2 = 0.74, RMSE= 15.52 kg DM ha-1day-1) versus Agmodel 2 (R2 = 0.58, RMSE= 13.74 kg DM ha-1day-1). This trained model was used on validation data from 2020, and the results were similar with better performance for Agmodel1 (R2 =0.70) versus Agmodel2 (R2=0.36). The improved spatial resolution of Sentinel 2 and the availability of red-edge bands showed improved results compared with previous work based on coarse resolution satellite imagery
Characterisation and monitoring of forest disturbances in Ireland using active microwave satellite platforms
Full text linkForests are one of the major carbon sinks that significantly contribute towards achieving
targets of the Kyoto Protocol, and its successors, in reducing greenhouse (GHG)
emissions. In order to contribute to regular National Inventory Reporting, and as part of
the on-going development of the Irish national GHG reporting system (CARBWARE),
improvements in characterisation of changes in forest carbon stocks have been
recommended to provide a comprehensive information flow into CARBWARE. The Irish
National Forest Inventory (NFI) is updated once every six years, thus there is a need for
an enhanced forest monitoring system to obtain annual forest updates to support
government agencies and forest management companies in their strategic decision making
and to comply with international GHG reporting standards. Sustainable forest
management is imperative to promote net carbon absorption from forests. Based on the
NFI data, Irish forests have removed or sequestered an average of 3.8 Mt of atmospheric
CO2 per year between 2007 and 2016. However, unmanaged and degraded forests become
a net emitter of carbon. Disturbances from human induced activities such as clear felling,
thinning and deforestation results in carbon emissions back into the atmosphere. Funded
by the Department of Agriculture, Food and the Marine (DAFM, Ireland), this PhD study
focuses on exploring the potential of data from L-band Synthetic Aperture Radar (SAR)
satellite based sensors for monitoring changes in the small stand forests of Ireland.
Historic data from ALOS PALSAR in the late 2000s and more recent data from ALOS-2
PALSAR-2 sensors have been used to map forest areas and characterise the different
disturbances observed within three different regions of Ireland. Forest mapping and
disturbance characterisation was achieved by combining the machine learning supervised
Random Forests (RF) and unsupervised Iterative Self-Organizing Data Analysis
(ISODATA) classification techniques. The lack of availability of ground truth data
supported use of this unsupervised approach which forms natural clusters based on their
multi-temporal signatures, with divergence statistics used to select the optimal number of
clusters to represent different forest classes. This approach to forest monitoring using SAR imagery has not been reported in the peer-review literature and is particularly beneficial
where there is a dearth of ground-based information. When applied to the forests, mapped
with an accuracy of up to 97% by RF, the ISODATA technique successfully identified
the unique multi-temporal pattern associated with clear-fells which exhibited a decrease
of 4 to 5 decibels (dB) between the images acquired before and after the event. The
clustering algorithm effectively highlighted the occurrence of other disturbance events
within forests with a decrease of 2±0.5dB between two consecutive years, as well as areas
of tree growth and afforestation.
A highlight of the work is the successful transferability of the algorithm, developed using
ALOS PALSAR, to ALOS-2 PALSAR-2 data thereby demonstrating the potential
continuity of annual forest monitoring. The higher spatial and radiometric resolutions of
ALOS-2 PALSAR-2 data have shown improvements in forest mapping compared to
ALOS PALSAR data. From mapping a minimum forest size of 1.8 ha with ALOS
PALSAR, a minimum area of 1.1 ha was achieved with the ALOS-2 PALSAR-2 images.
Moreover, even with some different backscatter characteristics of images acquired in
different seasons, similar signature patterns between the sensors were retrieved that helped
to define the cluster groups, thus demonstrating the robustness of the algorithm and its
successful transferability.
Having proven the potential to monitor forest disturbances, the results from both the
sensors were used to detect deforestation over the time period 2007-2016. Permanent
land-use changes pertaining to conversion of forests to agricultural lands and windfarms
were identified which are important with respect to forest monitoring and carbon reporting
in Ireland.
Overall, this work has presented a viable approach to support forest monitoring operations
in Ireland. By providing disturbance information from SAR, it can supplement projects
working with optical images which are generally limited by cloud cover, particularly in
parts of northern, western and upland Ireland. This approach adds value to ground based
forest monitoring by mapping distinct forests over large areas on an annual basis. This
study has demonstrated the ability to apply the algorithm to three different study areas,
with a vision to operationalise the algorithm on a national scale. The main limitations
experienced in this study were the lack of L-band SAR data availability and reference
datasets. With typically only one image acquired per year, and discrepancies and
omissions existing within reference datasets, understanding the behaviour of certain
cluster groups representing disturbances was challenging. However, this approach has
addressed some issues within the reference datasets, for example locating areas for which
a felling licence was granted but where trees were never cut, by providing detailed
systematic mapping of forests. Future satellites such as Tandem-L, SAOCOM-2A and 2B,
P-band BIOMASS mission and ALOS-4 PALSAR-3 may overcome the issue of limited
SAR image acquisitions provided more images per year are available, especially during
the summer months
Retrieval of grassland biophysical parameters using multitemporal optical and radar satellite data
No full textThe amount and quality of available biomass is a key factor for the sustainable livestock industry and agricultural management related decision making. Globally 31.5% of land cover is grassland while 80% of Ireland’s agricultural land is grassland. In Ireland, grasslands are intensively managed and provide the cheapest feed source for animals. This dissertation presents a detailed state of the art review of satellite remote sensing of grasslands, and the potential application of optical (Moderate–resolution Imaging Spectroradiometer (MODIS)) and radar (TerraSAR-X) time series imagery to estimate the grassland biomass at two study sites (Moorepark and Grange) in the Republic of Ireland using both statistical and state of the art machine learning algorithms. High quality weather data available from the on-site weather station was also used to calculate the Growing Degree Days (GDD) for Grange to determine the impact of ancillary data on biomass estimation. In situ and satellite data covering 12 years for the Moorepark and 6 years for the Grange study sites were used to predict grassland biomass using multiple linear regression, Neuro Fuzzy Inference Systems (ANFIS) models. The results demonstrate that a dense (8-day composite) MODIS image time series, along with high quality in situ data, can be used to retrieve grassland biomass with high performance (R2 = 0:86; p < 0:05, RMSE = 11.07 for Moorepark). The model for Grange was modified to evaluate the synergistic use of vegetation indices derived from remote sensing time series and accumulated GDD information. As GDD is strongly linked to the plant development, or phonological stage, an improvement in biomass estimation would be expected. It was observed that using the ANFIS model the biomass estimation accuracy increased from R2 = 0:76 (p < 0:05) to R2 = 0:81 (p < 0:05) and the root mean square error was reduced by 2.72%. The work on the application of optical remote sensing was further developed using a TerraSAR-X Staring Spotlight mode time series over the Moorepark study site to explore the extent to which very high resolution Synthetic Aperture Radar (SAR) data of interferometrically coherent paddocks can be exploited to retrieve grassland biophysical parameters. After filtering out the non-coherent plots it is demonstrated that interferometric coherence can be used to retrieve grassland biophysical parameters (i. e., height, biomass), and that it is possible to detect changes due to the grass growth, and grazing and mowing events, when the temporal baseline is short (11 days). However, it not possible to automatically uniquely identify the cause of these changes based only on the SAR backscatter and coherence, due to the ambiguity caused by tall grass laid down due to the wind. Overall, the work presented in this dissertation has demonstrated the potential of dense remote sensing and weather data time series to predict grassland biomass using machine-learning algorithms, where high quality ground data were used for training. At present a major limitation for national scale biomass retrieval is the lack of spatial and temporal ground samples, which can be partially resolved by minor modifications in the existing PastureBaseIreland database by adding the location and extent ofeach grassland paddock in the database. As far as remote sensing data requirements are concerned, MODIS is useful for large scale evaluation but due to its coarse resolution it is not possible to detect the variations within the fields and between the fields at the farm scale. However, this issue will be resolved in terms of spatial resolution by the Sentinel-2 mission, and when both satellites (Sentinel-2A and Sentinel-2B) are operational the revisit time will reduce to 5 days, which together with Landsat-8, should enable sufficient cloud-free data for operational biomass estimation at a national scale. The Synthetic Aperture Radar Interferometry (InSAR) approach is feasible if there are enough coherent interferometric pairs available, however this is difficult to achieve due to the temporal decorrelation of the signal. For repeat-pass InSAR over a vegetated area even an 11 days temporal baseline is too large. In order to achieve better coherence a very high resolution is required at the cost of spatial coverage, which limits its scope for use in an operational context at a national scale. Future InSAR missions with pair acquisition in Tandem mode will minimize the temporal decorrelation over vegetation areas for more focused studies. The proposed approach complements the current paradigm of Big Data in Earth Observation, and illustrates the feasibility of integrating data from multiple sources. In future, this framework can be used to build an operational decision support system for retrieval of grassland biophysical parameters based on data from long term planned optical missions (e. g., Landsat, Sentinel) that will ensure the continuity of data acquisition. Similarly, Spanish X-band PAZ and TerraSAR-X2 missions will ensure the continuity of TerraSAR-X and COSMO-SkyMed
- …
