Gareth Ireland

Examining the Capability of Supervised Machine Learning Classifiers in Extracting Flooded Areas from Landsat TM Imagery: A Case Study from a Mediterranean Flood

This study explored the capability of Support Vector Machines (SVMs) and regularised kernel Fisher’s discriminant analysis (rkFDA) machine learning supervised classifiers in extracting flooded area from optical Landsat TM imagery. The ability of both techniques was evaluated using a case study of a riverine flood event in 2010 in a heterogeneous Mediterranean region, for which TM imagery acquired shortly after the flood event was available. For the two classifiers, both linear and non-linear (kernel) versions were utilised in their implementation. The ability of the different classifiers to map the flooded area extent was assessed on the basis of classification accuracy assessment metrics. Results showed that rkFDA outperformed SVMs in terms of accurate flooded pixels detection, also producing fewer missed detections of the flooded area. Yet, SVMs showed less false flooded area detections. Overall, the non-linear rkFDA classification method was the more accurate of the two techniques (OA = 96.23%, K = 0.877). Both methods outperformed the standard Normalized Difference Water Index (NDWI) thresholding (OA = 94.63, K = 0.818) by roughly 0.06 K points. Although overall accuracy results for the rkFDA and SVMs classifications only showed a somewhat minor improvement on the overall accuracy exhibited by the NDWI thresholding, notably both classifiers considerably outperformed the thresholding algorithm in other specific accuracy measures (e.g. producer accuracy for the “not flooded” class was ~10.5% less accurate for the NDWI thresholding algorithm in comparison to the classifiers, and average per-class accuracy was ~5% less accurate than the machine learning models). This study provides evidence of the successful application of supervised machine learning for classifying flooded areas in Landsat imagery, where few studies so far exist in this direction. Considering that Landsat data is open access and has global coverage, the results of this study offers important information towards exploring the possibilities of the use of such data to map other significant flood events from space in an economically viable way.

Addressing the ability of a land biosphere model to predict key biophysical vegetation characterisation parameters with Global Sensitivity Analysis

Sensitivity Analysis (SA) of the SimSphere Soil Vegetation Atmosphere Transfer (SVAT) model has been performed in this study using a cutting edge and robust Global Sensitivity Analysis (GSA) approach, based on the use of the Gaussian Emulation Machine for Sensitivity Analysis (GEM-SA) tool. The sensitivity of the following model outputs was evaluated: the ambient CO2 concentration, the rate of CO2 uptake by the plant, the ambient O3 concentration, the flux of O3 from the air to the plant/soil boundary and the flux of O3 taken up by the plant alone. The most sensitive model inputs for the majority of outputs were: The Leaf Area Index (LAI), Fractional Vegetation Cover (Fr), Cuticle Resistance (CR) and Vegetation Height (VH). The influence of the external CO2 on the leaf and O3 concentration in the air as input parameters was also significant. Our study provides an important step forward in the global efforts towards SimSphere verification given the increasing interest in its use as an independent modelling or educational tool. Results of this study are also timely given the ongoing global efforts focused on deriving, at an operational level, spatio-temporal estimates of energy fluxes and soil moisture content using SimSphere synergistically with Earth Observation (EO) data. First study to examine the sensitivity of specific key outputs simulated from SimSphere.First study to examine the effect of simulation time on model sensitivity.Results extend our understanding of the SimSphere models structure and coherence.Results are potentially important for the development of operational EO products.

An appraisal of the accuracy of operational soil moisture estimates from SMOS MIRAS using validated in situ observations acquired in a Mediterranean environment

Acquiring information on the spatio-temporal variability of soil moisture is of key importance in extending our capability to understand the Earth system’s physical processes, and is also required in many practical applications. Earth observation (EO) provides a promising avenue to observe the distribution of soil moisture at different observational scales, with a number of products distributed at present operationally. Validation of such products at a range of climate and environmental conditions across continents is a fundamental step related to their practical use. Various in situ soil moisture ground observational networks have been established globally providing suitable data for evaluating the accuracy of EO-based soil moisture products. This study aimed at evaluating the accuracy of soil moisture estimates provided from the Soil Moisture and Ocean Salinity Mission (SMOS) global operational product at test sites from the REMEDHUS International Soil Moisture Network (ISMN) in Spain. For this purpose, validated observations from in situ ground observations acquired nearly concurrent to SMOS overpass were utilized. Overall, results showed a generally reasonable agreement between the SMOS product and the in situ soil moisture measurements in the 0–5 cm soil moisture layer (root mean square error (RMSE) = 0.116 m3 m−3). An improvement in product accuracy for the overall comparison was shown when days of high radio frequency interference were filtered out (RMSE = 0.110 m3 m−3). Seasonal analysis showed highest agreement during autumn, followed by summer, winter, and spring seasons. A systematic soil moisture underestimation was also found for the overall comparison and during the four seasons. Overall, the result provides supportive evidence of the potential value of this operational product for meso-scale studies and practical applications.

An appraisal of soil moisture operational estimates accuracy from SMOS MIRAS using validated in-situ observations acquired at a Mediterranean environment.

Acquiring information on the spatio-temporal variability of soil moisture is of key importance in extending our capability to understand the Earth system’s physical processes, and is also required in many practical applications. Earth observation (EO) provides a promising avenue to observe the distribution of soil moisture at different observational scales, with a number of products distributed at present operationally. Validation of such products at a range of climate and environmental conditions across continents is a fundamental step related to their practical use. Various in situ soil moisture ground observational networks have been established globally providing suitable data for evaluating the accuracy of EO-based soil moisture products. This study aimed at evaluating the accuracy of soil moisture estimates provided from the Soil Moisture and Ocean Salinity Mission (SMOS) global operational product at test sites from the REMEDHUS International Soil Moisture Network (ISMN) in Spain. For this purpose, validated observations from in situ ground observations acquired nearly concurrent to SMOS overpass were utilized. Overall, results showed a generally reasonable agreement between the SMOS product and the in situ soil moisture measurements in the 0–5 cm soil moisture layer (root mean square error (RMSE) = 0.116 m3 m−3). An improvement in product accuracy for the overall comparison was shown when days of high radio frequency interference were filtered out (RMSE = 0.110 m3 m−3). Seasonal analysis showed highest agreement during autumn, followed by summer, winter, and spring seasons. A systematic soil moisture underestimation was also found for the overall comparison and during the four seasons. Overall, the result provides supportive evidence of the potential value of this operational product for meso-scale studies and practical applications.

Operational evapotranspiration estimates from SEVIRI in support of sustainable water management

This study aimed at evaluating the accuracy of the evapotranspiration (ET) operational estimates from the Meteosat Second Generation (MSG) Spinning Enhanced Visible Infra-Red Imager (SEVIRI) at a range of selected ecosystems in Europe. For this purpose in-situ eddy covariance measurements were used, acquired from 7 selected experimental sites belonging to the CarboEurope ground observational network over 2 full years of observations (2010–2011). Appraisal of ET accuracy was also investigated with respect to land cover, season and each site(s) degree of heterogeneity, the latter being expressed by the fractional vegetation cover (FVC) operational product of SEVIRI. Results indicated a close agreement between the operational product’s ET estimates and the tower based in-situ ET measurements for all days of comparison, showing a satisfactory correlation (r of 0.709) with accuracies often comparable to previous analogous studies. For all land cover types, the grassland and cropland sites exhibited the closest agreement (r from 0.705 to 0.759). In terms of seasons the strongest correlations were observed during the summer and autumn (r of 0.714 & 0.685 respectively), and with FVC the highest correlation of 0.735 was observed for the class FVC 0.75-1 when compared against the observed values for the complete monitoring period. Our findings support the potential value of the SEVIRI ET product for regional to mesoscale studies and corroborate its credibility for usage in many practical applications. The latter is of particular importance for water limiting environments, such as those found in the Mediterranean basin, as accurate information on ET rates can provide tremendous support in sustainable water resource management as well as policy and decision making in those areas.

Performance Assessment of the SEVIRI Evapotranspiration Operational Product: Results Over Diverse Mediterranean Ecosystems

Evapotranspiration (ET) is an important variable in weather systems and hydrometeorological modeling. In this paper, an extensive validation was carried out on the spinning enhanced visible and infrared imager (SEVIRI) ET operational product, evaluating its accuracy at selected European sites. Validation was performed through comparisons with in-situ eddy covariance measurements belonging to the CarboEurope IP network. Comparisons were performed for selected cloud-free days with a satisfactory energy balance ratio in 2011. A total of nine sites covering six land covers were used in validating the ET retrieval accuracy from the operational product. A series of statistical metrics was computed to evaluate the agreement, which also included explored the variability of site characteristics and influence of land cover on ET performances. Overall, a good agreement was reported between the satellite-derived ET estimates and the ground measurements (d-index = 0.755, root mean square deviation (RMSD) = 0.107 mm h-1). A minor negative bias of -0.015 mm h-1 suggested only slight underestimation of the in-situ data. In terms of land cover, the highest agreement in ET was reported for the olive orchards and open shrubland sites (d-index = 0.893/0.867, RMSD = 0.041/0.050 mm h-1). A systematic ET underestimation by SEVIRI was found for all land cover types. Results of this study are largely in agreement to previous analogous validation studies of the product. Our findings support the potential value of the SEVIRI ET product for regional to mesoscale studies and practical applications. The latter is of particular importance for water limiting environments such as those found in the Mediterranean basin, as accurate information on ET rates can provide tremendous support in sustainable water resource management as well as policy and decision making.

Evaluation of the Soil Moisture Operational Estimates From SMOS in Europe: Results Over Diverse Ecosystems

This study presents the results of an extensive validation of the Soil Moisture and Ocean Salinity mission (SMOS) soil moisture operational product from selected European sites representative of a variety of climatic, environmental, biome, and seasonal conditions. SMOS soil moisture estimates were compared against corresponding in-situ measurements from the CarboEurope observational network. The agreement between the two datasets was evaluated on the basis of a series of statistical metrics. In addition, the effect of variability of site characteristics such as land cover, seasonality, and also that of the Radio Frequency Interference (RFI) effect on SMOS was explored. In overall, the SMOS soil moisture product estimates agreed reasonably well with near concurrent CarboEurope in-situ measurements acquired from the 0-5 cm soil moisture layer. Significant changes in the SMOS performance were observed with local adjustments, such as land cover and seasonal changes. Agreement was found to be higher over low vegetation cover and during the autumn season. The RFI contaminated pixels were filtered out from the pooled datasets, as well as from the seasonally discriminated datasets, which resulted in noticeably improved performances. This paper provides strong supportive evidence of the potential value of the SMOS soil moisture product for hydrometeorological and related studies.

A modernized version of a 1D soil vegetation atmosphere transfer model for improving its future use in land surface interactions studies

SimSphere is a land biosphere model that provides a mathematical representation of vertical views of the physical mechanisms controlling Earths energy and mass transfers in the soil/vegetation/atmosphere continuum. Herein, we present recent advancements introduced to SimSphere code, aiming at making its use more integrated to the automation of processes within High Performance Computing (HPC) that allows using the model at large scale. In particular, a new interface to the model is presented, so-called SimSphere-SOA which forms a command line land biosphere tool, a Web Service interface and a parameters verification facade that offers a standardised environment for specification execution and result retrieval of a typical model simulation based on Service Oriented Architecture (SOA). SimSphere-SOA library can now execute various simulations in parallel. This allows exploitation of the tool in a simple and efficient way in comparison to the currently distributed approach. In SimSphere-SOA, an Application Programming Interface (API) is also provided to execute simulations that can be publicly consumed. Finally this API is exported as a Web Service for remotely executing simulations through web based tools. This way a simulation by the model can be executed efficiently and subsequently the model simulation outputs may be used in any kind of relevant analysis required.The use of these new functionalities offered by SimSphere-SOA is also demonstrated using a real world simulation configuration file. The inclusion of those new functions in SimSphere are of considerable importance in the light of the models expanding use worldwide as an educational and research tool. Application Programming Interface (API) is now provided to execute SimSphere.A Service Oriented Architecture with Web Service interface added to SimSphere.SimSphere becomes now suitable for HPC use.Developments vital in its future use as standalone tool and for its synergy with EO data.

Chapter 5 – Spatiotemporal Estimates of Surface Soil Moisture from Space Using the Ts/VI Feature Space

Abstract Earth Observation (EO) has played an imperative role in extending our abilities for obtaining information on the spatio-temporal distribution of surface soil moisture (SSM). A wide range of techniques have been proposed for this purpose. Some of those techniques have based on the integration of satellite-derived estimates of Fractional Vegetation Cover (Fr) and Land Surface Temperature (Ts) in the form of a scatterplot domain, often combining land surface process model simulations. These techniques aim at combining the horizontal coverage and spectral resolution of EO imagery with the vertical coverage and fine temporal continuity of the process models. Herein one such technique - named the “triangle” - has been implemented with EO datasets from both the AATSR and ASTER sensors together with SimSphere land surface model. Validation of the derived SMC maps was undertaken in different sites in Europe representing a variety of climatic, topographic and environmental conditions, for which validated in-situ observations from diverse operational ground observational networks were available. Results indicated a good agreement between the in-situ and both “triangle” schemes for the estimation of SMC (ASTER R – 0.561/AATSR R – 0.844), with the AATSR results again outperforming the ASTER, comparable to previous studies implemented using different types of EO data. Comparisons of the derived SMC maps regionally against other satellite-derived products also showed largely an explainable distribution of SMC in relation to surface heterogeneity. Our results provide strong supportive evidence for the potential value of the “triangle” inversion modelling technique to accurately derive estimates of SMC, and are important steps as well towards efforts focusing on operational implementation of this approach.

PROgRESSIon—Investigating the Prototyping of Operational Estimation of Energy Fluxes and Soil Moisture Content Using a Variant of the “Triangle” Inversion Methodology

Accurately estimating the spatio-temporal distribution of energy, mass and momentum at the surface–atmosphere interface can help develop a better understanding of the complex interactions of the Earth system. By linking deterministic land surface process model, such as SimSphere, to the spatialised information provided by Earth observation (EO) data, a more powerful synergistic avenue can be developed to take advantage of the temporal and spatial benefits of both modelling and EO-based approaches. The “triangle” utilises the distribution of land surface temperature (LST) and vegetation index (VI) formed by a satellite-derived scatterplot, linked with SimSphere under a full range of vegetation cover and soil moisture, to derive spatial estimates of energy fluxes and soil moisture content (SMC). To this end, the objective of this study was to implement the “triangle” technique using Advanced Along-Track Scanning Radiometer (AATSR) satellite data products to derive and subsequently validate spatially explicit maps of land surface heat fluxes and SM for different ecosystems in Europe. The “triangle”-derived estimations of soil moisture exhibited a minor overestimation of the in-situ observations, and an average error of 0.097 vol vol−1. In overall, results were comparable to those of previous validation studies of the “triangle” implementation. Results for the LE and H fluxes were within the accuracy range of 50 Wm−2, with root mean square difference (RMSD) of 41.15 Wm−2 and 44.37 Wm−2, respectively. Furthermore, there was a good agreement between the “triangle”-derived and in-situ observed instantaneous LE and H fluxes, exhibited by high R values (0.88 and 0.69, respectively). Our study is one of the few studies validating the “triangle” over different ecosystems in Europe. It is a significant step forward in supporting the operational development of this method using remote sensing data in deriving key land surface parameters on a global scale.