Dissertations / Theses on the topic 'Multiscale remote sensing'

La humedad del suelo es la variable que regula los intercambios de agua, energía, y carbono entre la tierra y la atmósfera. Mediciones precisas de humedad son necesarias para una gestión sostenible de los recursos hídricos, para mejorar las predicciones meteorológicas y climáticas, y para la detección y monitorización de sequías e inundaciones. Esta tesis se centra en la medición de la humedad superficial de la Tierra desde el espacio, a escalas global y regional. Estudios teóricos y experimentales han demostrado que la teledetección pasiva de microondas en banda L es optima para la medición de humedad del suelo, debido a que la atmósfera es transparente a estas frecuencias, y a la relación directa de la emisividad del suelo con su contenido de agua. Sin embargo, el uso de la teledetección pasiva en banda L ha sido cuestionado en las últimas décadas, pues para conseguir la resolución temporal y espacial requeridas, un radiómetro convencional necesitaría una gran antena rotatoria, difícil de implementar en un satélite. Actualmente, hay tres principales propuestas para abordar este problema: (i) el uso de un radiómetro de apertura sintética, que es la solución implementada en la misión Soil Moisture and Ocean Salinity (SMOS) de la ESA, en órbita desde noviembre del 2009; (ii) el uso de un radiómetro ligero de grandes dimensiones y un rádar operando en banda L, que es la solución que ha adoptado la misión Soil Moisture Active Passive (SMAP) de la NASA, con lanzamiento previsto en 2014; (iii) el desarrollo de técnicas de desagregación de píxel que permitan mejorar la resolución espacial de las observaciones. La primera parte de la tesis se centra en el estudio del algoritmo de recuperación de humedad del suelo a partir de datos SMOS, que es esencial para obtener estimaciones de humedad con alta precisión. Se analizan diferentes configuraciones con datos simulados, considerando (i) la opción de añadir información a priori de los parámetros que dominan la emisión del suelo en banda L —humedad, rugosidad, temperatura del suelo, albedo y opacidad de la vegetación— con diferentes incertidumbres asociadas, y (ii) el uso de la polarización vertical y horizontal por separado, o del primer parámetro de Stokes. Se propone una configuración de recuperación de humedad óptima para SMOS. La resolución espacial de los radiómetros de SMOS y SMAP (40-50 km) es adecuada para aplicaciones globales, pero limita la aplicación de los datos en estudios regionales, donde se requiere una resolución de 1-10 km. La segunda parte de esta tesis contiene tres novedosas propuestas de mejora de resolución espacial de estos datos: • Se ha desarrollado un algoritmo basado en la deconvolución de los datos SMOS que permite mejorar la resolución espacial de las medidas. Los resultados de su aplicación a datos simulados y a datos obtenidos con un radiómetro aerotransportado muestran que es posible mejorar el producto de resolución espacial y resolución radiométrica de los datos. • Se presenta un algoritmo para mejorar la resolución espacial de las estimaciones de humedad de SMOS utilizando datos MODIS en el visible/infrarrojo. Los resultados de su aplicación a algunas de las primeras imágenes de SMOS indican que la variabilidad espacial de la humedad del suelo se puede capturar a 32, 16 y 8 km. • Un algoritmo basado en detección de cambios para combinar los datos del radiómetro y el rádar de SMAP en un producto de humedad a 10 km ha sido desarrollado y validado utilizando datos simulados y datos experimentales aerotransportados. Este trabajo se ha desarrollado en el marco de las actividades preparatorias de SMOS y SMAP, los dos primeros satélites dedicados a la monitorización de la variación temporal y espacial de la humedad de la Tierra. Los resultados presentados contribuyen a la obtención de estimaciones de humedad del suelo con la precisión y la resolución espacial necesarias para un mejor conocimiento del ciclo del agua y una mejor gestión de los recursos hídricos.
Soil moisture is a key state variable of the Earth's system; it is the main variable that links the Earth's water, energy and carbon cycles. Accurate observations of the Earth's changing soil moisture are needed to achieve sustainable land and water management, and to enhance weather and climate forecasting skill, flood prediction and drought monitoring. This Thesis focuses on measuring the Earth's surface soil moisture from space at global and regional scales. Theoretical and experimental studies have proven that L-band passive remote sensing is optimal for soil moisture sensing due to its all-weather capabilities and the direct relationship between soil emissivity and soil water content under most vegetation covers. However, achieving a temporal and spatial resolution that could satisfy land applications has been a challenge to passive microwave remote sensing in the last decades, since real aperture radiometers would need a large rotating antenna, which is difficult to implement on a spacecraft. Currently, there are three main approaches to solving this problem: (i) the use of an L-band synthetic aperture radiometer, which is the solution implemented in the ESA Soil Moisture and Ocean Salinity (SMOS) mission, launched in November 2009; (ii) the use of a large lightweight radiometer and a radar operating at L-band, which is the solution adopted by the NASA Soil Moisture Active Passive (SMAP) mission, scheduled for launch in 2014; (iii) the development of pixel disaggregation techniques that could enhance the spatial resolution of the radiometric observations. The first part of this work focuses on the analysis of the SMOS soil moisture inversion algorithm, which is crucial to retrieve accurate soil moisture estimations from SMOS measurements. Different retrieval configurations have been examined using simulated SMOS data, considering (i) the option of adding a priori information from parameters dominating the land emission at L-band —soil moisture, roughness, and temperature, vegetation albedo and opacity— with different associated uncertainties and (ii) the use of vertical and horizontal polarizations separately, or the first Stokes parameter. An optimal retrieval configuration for SMOS is suggested. The spatial resolution of SMOS and SMAP radiometers (~ 40-50 km) is adequate for global applications, but is a limiting factor to its application in regional studies, where a resolution of 1-10 km is needed. The second part of this Thesis contains three novel downscaling approaches for SMOS and SMAP: • A deconvolution scheme for the improvement of the spatial resolution of SMOS observations has been developed, and results of its application to simulated SMOS data and airborne field experimental data show that it is feasible to improve the product of the spatial resolution and the radiometric sensitivity of the observations by 49% over land pixels and by 30% over sea pixels. • A downscaling algorithm for improving the spatial resolution of SMOS-derived soil moisture estimates using higher resolution MODIS visible/infrared data is presented. Results of its application to some of the first SMOS images show the spatial variability of SMOS-derived soil moisture observations is effectively captured at the spatial resolutions of 32, 16, and 8 km. • A change detection approach for combining SMAP radar and radiometer observations into a 10 km soil moisture product has been developed and validated using SMAP-like observations and airborne field experimental data. This work has been developed within the preparatory activities of SMOS and SMAP, the two first-ever satellites dedicated to monitoring the temporal and spatial variation on the Earth's soil moisture. The results presented contribute to get the most out of these vital observations, that will further our understanding of the Earth's water cycle, and will lead to a better water resources management.

This study investigated the use of optical remote sensing for estimating leaf and canopy scale light use efficiency (LUE) and carbon exchange. In addition, a new leaf level model capable of predicting dynamic changes in apparent reflectance due to chlorophyll fluorescence was developed. A leaf level study was conducted to assess the applicability of passive remote sensing as a tool to measure the reduction, and the subsequent recovery, of photosynthetic efficiency during the weeks following transplantation. Spectral data were collected on newly planted saplings for a period of 8 weeks, as well as gas exchange measurements of LUE and PAM fluorescence measurements. A set of spectral indices, including the Photochemical Reflectance Index (PRI), were calculated from the reflectance measurements. A marked depression in photosynthetic rate occurred in the weeks after outplanting followed by a gradual increase, with recovery occurring in the later stages of the experimental period. As with photosynthetic rate, there was a marked trend in PRI values over the study period but no trend was observed in chlorophyll based indices. The study demonstrated that hyperspectral remote sensing has the potential to be a useful tool in the detection and monitoring of the dynamic effects of transplant shock. Relationships between hyperspectral reflectance indices, airborne carbon exchange measurements and satellite observations of ground cover were then explored across a heterogeneous Arctic landscape. Measurements were collected during August 2008, using the University of Edinburgh’s research aircraft, from an Arctic forest tundra zone in northern Finland as part of the Arctic Biosphere Atmosphere Coupling at Multiple Scales (ABACUS) study. Surface fluxes of CO2 were calculated using the eddy covariance method from airborne data that were collected from the same platform as hyperspectral reflectance measurements. Airborne CO2 fluxes were compared to MODIS vegetation indices. In addition, LUE was estimated from airborne flux data and compared to airborne measurements of PRI. There were no significant relationships between MODIS vegetation indices and airborne flux observations. There were weak to moderate (R2 = 0.4 in both cases) correlations between PRI and LUE and between PRI and incident radiation. A new coupled physiological radiative transfer model that predicts changes in the apparent reflectance of a leaf, due to chlorophyll fluorescence, was developed. The model relates a physically observable quantity, chlorophyll fluorescence, to the sub leaf level processes that cause the emission. An understanding of the dynamics of the processes that control fluorescence emission on multiple timescales should aid in the interpretation of this complex signal. A Markov Chain Monte Carlo (MCMC) algorithm was used to optimise biochemical model parameters by fitting model simulations of transient chlorophyll fluorescence to measured reflectance spectra. The model was then validated against an independent data set. The model was developed as a precursor to a full canopy scheme. To scale to the canopy and to use the model on trans-seasonal time scales, the effects of temperature and photoinhibition on the model biochemistry needs to be taken into account, and a full canopy radiative transfer scheme, such as FluorMOD, must be developed.

The health of natural vegetation communities is of concern due to observed changes in the climatic-hydrological regime and land cover changes particularly in arid and semiarid regions. Monitoring vegetation at multi temporal and spatial scales can be the most informative approach for detecting change and inferring causal agents of change and remediation strategies. Riparian communities are tightly linked to annual stream hydrology, ground water elevations and sediment transport. These processes are subject to varying magnitudes of disturbance overtime and are candidates for multi-scale monitoring. My first research objective focused on the response of vegetation in the Upper San Pedro River, Arizona, to reduced base flows and climate change. I addressed the correlation between riparian vegetation and hydro-climate variables during the last three decades in one of the remaining undammed rivers in the southwestern U.S. Its riparian forest is threatened by the diminishing base flows, attributed by different studies either to increases in evapotranspiration (ET) due to conversion of grasslands to mesquite shrublands in the adjacent uplands, or to increased regional groundwater pumping to serve growing populations in surrounding urban areas and or to some interactions of those causes. Landsat 5 imagery was acquired for pre- monsoon period, when riparian trees had leafed out but before the arrival of summer monsoon rains in July. The result has showed Normalized Difference Vegetation Index (NDVI) values from both Landsat and Moderate Resolution Imaging Spectrometer (MODIS) had significant decreases which positively correlated to river flows, which decreased over the study period, and negatively correlated with air temperatures, which have increased by about 1.4°C from 1904 to the present. The predictions from other studies that decreased river flows could negatively impact the riparian forest were supported by this study. The pre-monsoon Normalized Different Vegetation Index (NDVI) average values in the adjacent uplands also decreased over thirty years and were correlated with the previous year's annual precipitation. Hence an increase in ET in the uplands did not appear to be responsible for the decrease in river flows in this study, leaving increased regional groundwater pumping as a feasible alternative explanation for decreased flows and deterioration of the riparian forest. The second research objective was to develop a new method of classification using very high-resolution aerial photo to map riparian vegetation at the species level in the Colorado River Ecosystem, Grand Canyon area, Arizona. Ground surveys have showed an obvious trend in which non-native saltcedar (Tamarix spp.) has replaced native vegetation over time. Our goal was to develop a quantitative mapping procedure to detect changes in vegetation as the ecosystem continues to respond to hydrological and climate changes. Vegetation mapping for the Colorado River Ecosystem needed an updated database map of the area covered by riparian vegetation and an indicator of species composition in the river corridor. The objective of this research was to generate a new riparian vegetation map at species level using a supervised image classification technique for the purpose of patch and landscape change detection. A new classification approach using multispectral images allowed us to successfully identify and map riparian species coverage the over whole Colorado River Ecosystem, Grand Canyon area. The new map was an improvement over the initial 2002 map since it reduced fragmentation from mixed riparian vegetation areas. The most dominant tree species in the study areas is saltcedar (Tamarix spp.). The overall accuracy is 93.48% and the kappa coefficient is 0.88. The reference initial inventory map was created using 2002 images to compare and detect changes through 2009. The third objective of my research focused on using multiplatform of remote sensing and ground calibration to estimate the effects of vegetation, land use patterns and water cycles. Climate change, hydrological and human uses are also leading to riparian, upland, grassland and crop vegetation changes at a variety of temporal and spatial scales, particularly in the arid and semi-arid ecosystems, which are more sensitive to changes in water availability than humid ecosystems. The objectives of these studies from the last three articles were to evaluate the effect of water balance on vegetation indices in different plant communities based on relevant spatial and temporal scales. The new methodology of estimating water requirements using remote sensing data and ground calibration with flux tower data has been successfully tested at a variety sites, a sparse desert shrub environment as well as mixed riparian and cropland systems and upland vegetation in the arid and semi-arid regions. The main finding form these studies is that vegetation-index methods have to be calibrated with ground data for each new ecosystem but once calibrated they can accurately scale ET over wide areas and long time spans.

The objective of this study was to investigate the application of multiscale satellite remote sensing data for assessment of land cover change in the rural-urban fringe. Inherent in this assessment process was the interpretation of multispectral data collected by several medium resolution satellite systems and evaluation of the quality of the resulting change information. Each dataset was acquired for a single date and classified at two levels of detail using standard classification algorithms. The optimum classification approach for each date was identified and the changes in land cover evaluated in several ways. The contribution of spatial and thematic errors and their propagation through the analysis process was investigated.Data for this research were acquired over an area approximately 4.5 km square located in the southern metropolitan area of Perth, Western Australia. At the time of the initial data acquisition in 1972 the area was predominantly rural and comprised mostly dense pine plantations, however by the final stages of data acquisition in 1991, the area was almost completely given over to urban residential land use. Changes were interpreted from classified Landsat Multispectral Scanner (MSS), Landsat Thematic Mapper (TM) and SPOT (System Pour l'Observation de la Terre) High Resolution Visible (HRV) multispectral data, and were compared to reference maps compiled from medium scale aerial photographs. The geometric properties of high resolution panchromatic IRS1-D data were also evaluated to test the geometric potential of high resolution satellite data.Supervised and unsupervised classification algorithms were used for derivation of land cover maps from each multispectral dataset at two levels of detail. Data were classified onto four general levels at the broadest (Level I) classification, and into nine levels at the finest (Level II) classification. The Kappa statistic and its variance were used to determine the optimum classification approach for each dataset and at each level of detail. No significant differences were observed between classification techniques at Level I, however at Level II the supervised classification approach produced significantly better results for the Landsat TM and SPOT HRV data. Classification at the more general Level I did not produce substantially higher classification rates compared to the same data at Level II. Additionally, higher spatial resolution data did not provide increased accuracy, however this was due mainly to a much greater complexity of land covers present at the time the higher resolution Landsat TM and SPOT HRV data were recorded.Land cover changes were assessed separately at Level I for all datasets, and also between Landsat TM and SPOT HRV data at Level II. Integrated multiscale assessment of land cover change was undertaken using classified Landsat MSS data at Level I and Landsat TM data at Level 11. This enabled the continuity of change to be established across classification levels and sensor systems, even though there were variations in the level of detail extracted from each image.The sources of spatial and thematic errors in the data were investigated and their effects on change assessment analysed. The evaluation of high resolution panchromatic satellite data emphasised the contribution to the analysis of spatial errors contained within the reference data. The multiscale data also indicated that combined propagation of spatial and thematic errors requires investigation using appropriate simulation modelling to establish the influence of data uncertainty on classification and change assessment results.This research provides useful results for demonstrating a process for the integration of information derived from remotely sensed data at different measurement scales. Availability of data from an increasing range of remote sensing platforms and uncertainty of long term data availability emphasises the need to develop flexible interpretation and analysis approaches. This research adds value to the existing data archive by demonstrating how historical data may be integrated regardless of the spectral and spatial characteristics of the sensors.

Thesis (Ph. D.)--Massachusetts Institute of Technology, Dept. of Electrical Engineering and Computer Science, 1995.
Vita.
Includes bibliographical references (p. 287-298).
by Paul Werner Fieguth.
Ph.D.

The objective of this study was to investigate the application of multiscale satellite remote sensing data for assessment of land cover change in the rural-urban fringe. Inherent in this assessment process was the interpretation of multispectral data collected by several medium resolution satellite systems and evaluation of the quality of the resulting change information. Each dataset was acquired for a single date and classified at two levels of detail using standard classification algorithms. The optimum classification approach for each date was identified and the changes in land cover evaluated in several ways. The contribution of spatial and thematic errors and their propagation through the analysis process was investigated.Data for this research were acquired over an area approximately 4.5 km square located in the southern metropolitan area of Perth, Western Australia. At the time of the initial data acquisition in 1972 the area was predominantly rural and comprised mostly dense pine plantations, however by the final stages of data acquisition in 1991, the area was almost completely given over to urban residential land use. Changes were interpreted from classified Landsat Multispectral Scanner (MSS), Landsat Thematic Mapper (TM) and SPOT (System Pour l'Observation de la Terre) High Resolution Visible (HRV) multispectral data, and were compared to reference maps compiled from medium scale aerial photographs. The geometric properties of high resolution panchromatic IRS1-D data were also evaluated to test the geometric potential of high resolution satellite data.Supervised and unsupervised classification algorithms were used for derivation of land cover maps from each multispectral dataset at two levels of detail. Data were classified onto four general levels at the broadest (Level I) classification, and into nine levels at the finest (Level II) classification. The ++
Kappa statistic and its variance were used to determine the optimum classification approach for each dataset and at each level of detail. No significant differences were observed between classification techniques at Level I, however at Level II the supervised classification approach produced significantly better results for the Landsat TM and SPOT HRV data. Classification at the more general Level I did not produce substantially higher classification rates compared to the same data at Level II. Additionally, higher spatial resolution data did not provide increased accuracy, however this was due mainly to a much greater complexity of land covers present at the time the higher resolution Landsat TM and SPOT HRV data were recorded.Land cover changes were assessed separately at Level I for all datasets, and also between Landsat TM and SPOT HRV data at Level II. Integrated multiscale assessment of land cover change was undertaken using classified Landsat MSS data at Level I and Landsat TM data at Level 11. This enabled the continuity of change to be established across classification levels and sensor systems, even though there were variations in the level of detail extracted from each image.The sources of spatial and thematic errors in the data were investigated and their effects on change assessment analysed. The evaluation of high resolution panchromatic satellite data emphasised the contribution to the analysis of spatial errors contained within the reference data. The multiscale data also indicated that combined propagation of spatial and thematic errors requires investigation using appropriate simulation modelling to establish the influence of data uncertainty on classification and change assessment results.This research provides useful results for demonstrating a process for the integration of information derived from remotely sensed data at different measurement ++
scales. Availability of data from an increasing range of remote sensing platforms and uncertainty of long term data availability emphasises the need to develop flexible interpretation and analysis approaches. This research adds value to the existing data archive by demonstrating how historical data may be integrated regardless of the spectral and spatial characteristics of the sensors.

The subtropical forests located along South Africa’s Wild Coast region, declared as one of the biodiversity hotspots, provide benefits to the local and national economy. However, there is evidence of increased pressure exerted on the forests by growing population and reduced income from activities not related to forest products. The ability of remote sensing to quantify subtropical forest changes over time, perform species discrimination (using field spectroscopy) and integrating field spectral and multispectral data were all assessed in this study. Investigations were conducted at pixel, leaf and sub-pixel levels. Both per-pixel and sub-pixel classification methods were used for improved forest characterisation. Using SPOT 6 imagery for 2013, the study determined the best classification algorithm for mapping sub-tropical forest and other land cover types to be the maximum likelihood classifier. Maximum likelihood outperformed minimum distance, spectral angle mapper and spectral information divergence algorithms, based on overall accuracy and Kappa coefficient values. Forest change analysis was made based on spectral measurements made at top of the atmosphere (TOC) level. When applied to the 2005 and 2009 SPOT 5 images, subtropical forest changes between 2005-2009 and 2009-2013 were quantified. A temporal analysis of forest cover trends in the periods 2005-2009 and 2009-2013 identified a decreasing trend of -3648.42 and -946.98 ha respectively, which translated to 7.81 percent and 2.20 percent decrease. Although there is evidence of a trend towards decreased rates of forest loss, more conservation efforts are required to protect the Wild Coast ecosystem. Using field spectral measurements data, the hierarchical method (comprising One-way ANOVA with Bonferroni correction, Classification and Regression Trees (CART) and Jeffries Matusita method) successfully selected optimal wavelengths for species discrimination at leaf level. Only 17 out of 2150 wavelengths were identified, thereby reducing the complexities related to data dimensionality. The optimal 17 wavelength bands were noted in the visible (438, 442, 512 and 695 nm), near infrared (724, 729, 750, 758, 856, 936, 1179, 1507 and 1673 nm) and mid-infrared (2220, 2465, 2469 and 2482 nm) portions of the electromagnetic spectrum. The Jeffries-Matusita (JM) distance method confirmed the separability of the selected wavelength bands. Using these 17 wavelengths, linear discriminant analysis (LDA) classified subtropical species at leaf level more accurately than partial least squares discriminant analysis (PLSDA) and random forest (RF). In addition, the study integrated field-collected canopy spectral and multispectral data to discriminate proportions of semi-deciduous and evergreen subtropical forests at sub-pixel level. By using the 2013 land cover (using MLC) to mask non-forested portions before sub-pixel classification (using MTMF), the proportional maps were a product of two classifiers. The proportional maps show higher proportions of evergreen forests along the coast while semi-deciduous subtropical forest species were mainly on inland parts of the Wild Coast. These maps had high accuracy, thereby proving the ability of an integration of field spectral and multispectral data in mapping semi-deciduous and evergreen forest species. Overall, the study has demonstrated the importance of the MLC and LDA and served to integrate field spectral and multispectral data in subtropical forest characterisation at both leaf and top-of-atmosphere levels. The success of both the MLC and LDA further highlighted how essential parametric classifiers are in remote sensing forestry applications. Main subtropical characteristics highlighted in this study were species discrimination at leaf level, quantifying forest change at pixel level and discriminating semi-deciduous and evergreen forests at sub-pixel level.

Concern for environmental management of our natural resources is most often focused on the human impacts upon these resources. Minor stresses on surface materials in sensitive desert landscapes can greatly increase the rate and character of erosion. The National Training Center, Ft. Irwin, located in the middle of the Mojave Desert, California, provides a study area of intense off-road vehicle (ORV) activity spanning a 50-year period. This study documents a case of concentrated ORV activity on sensitive desert environments, and the resulting environmental impacts. Geomorphic surfaces from two study sites within the Ft. Irwin area were mapped from 1:28,400 scale black and white aerial photographs taken in 1947. Surface disruption attributed to military activity was then mapped for the same areas from 1993, 1:12,000, black and white aerial photographs. Several field checks were conducted to verify this mapping. Images created from SPOT panchromatic and Landsat Thematic Mapper (TM) multispectral data acquired during the spring of 1987 and 1993 were analyzed to assess both the extent of disrupted surfaces and the surface geomorphology discernable from satellite data. Classified and merged images were then created from these data and demonstrate the capabilities of satellite data to aid in the delineation of disrupted geomorphic surfaces. Correlations were also established between highly disrupted surfaces and soil surface conditions on selected geomorphic surfaces. Disruption maps produced from the air photos indicate that the amount of disrupted surfaces within the study sites grew from a combined total of 1.3 km² in 1947 to 33.4 km² by 1993. A combination of 6 bands of Landsat TM data with a seventh band of SPOT panchromatic data yielded a product that delineated broad geomorphic surfaces that closely correlate with those mapped from the aerial photography. An error matrix between these two products resulted in an overall accuracy of 83.36% and a Kappa Index of Agreement of 77.28%. A 15-class unsupervised classification of the SPOT panchromatic data produced the representation of the extent and levels of disruption present in the study areas that closely matched field observations. Field sampling of soil strength and clay/silt percentages on disturbed and undisturbed surfaces reveals that these arid land surfaces react to intense ORV activity by becoming more compact and exhibiting higher percentages of clays and silts.

Les imatges de teledetecció de la superfície oceànica proporcionen una vista sinòptica de la complexa geometria de la circulació oceànica, dominada per la variabilitat de mesoescala. Estructures com filaments i vòrtex són presents en els diferents escalars advectats pel flux oceànic. L’origen més probable d’aquestes estructures és el caràcter turbulent dels corrents, aquestes estructures són persistents amb el temps i compatibles amb la dinàmica mesoscalar oceànica. A escales espacials de quilòmetres o més, la turbulència és principalment 2D, i una complexa geometria, plena de filaments i remolins de mides diferents, emergeix en les imatges superficials de teledetecció de concentració de clorofil·la-a, salinitat superficial, així com en altres escalars més coneguts com són la temperatura superficial i la topografia dinàmica. L’objectiu d’aquesta tesi és explorar i aplicar metodologies de mapatge que permeten millorar la qualitat de mapes de teledetecció oceànica en general, i en particular de la salinitat superficial del mar (SSS). Les diferents metodologies emprades en aquesta tesi han estat aplicades amb l’objectiu específic de millorar els mapes de teledetecció de salinitat superficial del mar proveïts per la missió SMOS de l’Agència Espaial Europea. SMOS és el primer satèl·lit capaç de mesurar la humitat del sol i salinitat oceànica des de l’espai a escala global. La primera part d’aquesta tesi se centra a analitzar les característiques dels productes de nivell 2 (L2) de salinitat de SMOS i produir mapes de nivell 3 (L3) de salinitat utilitzant aproximacions clàssiques: millora del filtratge, mitjana ponderada i Interpolació Òptima. En el curs de la nostra recerca obtenim un conjunt de recomanacions de com processar les dades de SMOS començant des del nivell L2. Aquesta tesi també presenta una nova tècnica de fusió de dades que permet explotar les estructures turbulentes comunes entre diferents variables oceàniques, representant un pas endavant en la cadena de processat per generar mapes de nivell 4 (L4). Aquesta tècnica de fusió es basa teòricament en les propietats geomètriques dels traçadors advectats per la dinàmica oceànica (Turiel et al., 2005a). Degut a l’efecte de forta cissalla als fluits turbulents, l’estructura espacial d’un traçador oceànic hereta algunes propietats del flux subjacent, i en particular el seu arranjament geomètric. Com a conseqüència, les diferents variables oceàniques mostren propietats d’escala similars a la cascada d’energia turbulenta (Seuront and Schmitt, 2005; Nieves et al., 2007; Nieves and Turiel, 2009; Isern-Fontanet et al., 2007). El mètode de fusió agafa un senyal de menor qualitat (afectat per soroll, forats de dades i/o de resolució més baixa) i en millora la seva qualitat. A més d’això, el mètode de fusió és capaç d’extrapolar les dades de forma geofísicament coherent. Aquesta millora del senyal s’aconsegueix utilitzant una altra variable oceànica adquirida amb major qualitat, cobertura espacial més gran i/o millor resolució. Un punt clau d’aquesta aproximació és la suposició de l’existència d’una estructura multifractal de les imatges de teledetecció oceànica (Lovejoy et al., 2001b), i que les línies de singularitat de les diferents variables de l’oceà coincideixen. Sota aquestes premises, els gradients de les dues variables a fusionar estan relacionats per una matriu suau. Com a primera i simple aproximació, s’assumeix que aquesta matriu és proporcional a la identitat; això porta a un esquema de regressió lineal local. Aquesta tesi mostra que aquesta aproximació senzilla permet reduir l’error i millorar la cobertura del producte de nivell 4 resultant. D’altra banda, s’obté informació sobre la relació estadística entre les dues variables fusionades, ja que la dependència funcional entre elles es determina per cada punt de la imatge.
Remote sensing imagery of the ocean surface provides a synoptic view of the complex geometry of ocean circulation, which is dominated by mesoscale variability. The signature of filaments and vortices is present in different ocean scalars advected by the oceanic flow. The most probable origin of the observed structures is the turbulent character of ocean currents, and those signatures are persistent over time scales compatible with ocean mesoscale dynamics. At spatial scales of kilometers or more, turbulence is mainly 2D, and a complex geometry, full of filaments and eddies of different sizes, emerges in remote sensing images of surface chlorophyll-a concentration and surface salinity, as well as in other scalars acquired with higher quality such as surface temperature and absolute dynamic topography. The aim of this thesis is to explore and apply mapping methodologies to improve the quality of remote sensing maps in general, but focusing in the case of remotely sensed sea surface salinity (SSS) data. The different methodologies studied in this thesis have been applied with the specific goal of improving surface salinity maps generated from data acquired by the European Space Agency's mission SMOS, the first satellite able to measure soil moisture and ocean salinity from space at a global scale. The first part of this thesis will introduce the characteristics of the operational SMOS Level 2 (L2) SSS products and the classical approaches to produce the best possible SSS maps at Level 3 (L3), namely data filtering, weighted average and Optimal Interpolation. In the course of our research we will obtain a set of recommendations about how to process SMOS data starting from L2 data. A fusion technique designed to exploit the common turbulent signatures between different ocean variables is also explored in this thesis, in what represents a step forward from L3 to Level 4 (L4). This fusion technique is theoretically based on the geometrical properties of advected tracers. Due to the effect of the strong shear in turbulent flows, the spatial structure of tracers inherit some properties of the underlying flow and, in particular, its geometrical arrangement. As a consequence, different ocean variables exhibit scaling properties, similar to the turbulent energy cascade. The fusion method takes a signal affected by noise, data gaps and/or low resolution, and improves it in a geophysically meaningful way. This signal improvement is achieved by using an appropriate data, which is another ocean variable acquired with higher quality, greater spatial coverage and/or finer resolution. A key point in this approach is the assumption of the existence of a multifractal structure in ocean images, and that singularity lines of the different ocean variables coincide. Under these assumptions, the horizontal gradients of both variables, signal and template, can be related by a smooth matrix. The first, simplest approach to exploit such an hypothesis assumes that the relating matrix is proportional to the identity, leading to a local regression scheme. As shown in the thesis, this simple approach allows reducing the error and improving the coverage of the resulting Level 4 product; Moreover, information about the statistical relationship between the two fields is obtained since the functional dependence between signal and template is determined at each point.

This research presents models for the analysis of textural and contextual information content of multiscale remote sensing to select an appropriate scale for the correct interpretation and mapping of heterogeneous urban land cover types. Spatial complexity measures such as the fractal model and the Moran’s I index of spatial autocorrelation were applied for addressing the issue of scale, while fuzzy set theory was applied for mapping heterogeneous urban land cover types. Three local government areas (e.g. the City of Perth, the City of Melville and the City of Armadale) of the Perth metropolitan area were selected, as the dominant land covers of these areas are representative to the whole metropolitan area, for the analysis of spatial complexity and the mapping of complex land covers. Characterisation of spatial complexity of the study areas computed from SPOT, Landsat-7 ETM+, and Landsat MSS was used for assessing the appropriateness of a scale for urban analysis. Associated with this outcome, the effect of spectral resolution and land cover heterogeneity on spatial complexity, the performance of fractal measurement algorithms and the relationship between the fractal dimension and Moran’s I were identified. A fuzzy supervised approach of the fuzzy c-means algorithm was used to generate fuzzy memberships of the selected bands of a Landsat-7 ETM+ scene based on the highest spectral separability among different urban land covers (e.g. forest, grassland, urban and dense urban) as determined by a transformed divergence analysis. Fuzzy land cover maps resulting from the application of fuzzy operators (e.g. maximum, minimum, algebraic sum, algebraic product and gamma operators) were evaluated against fuzzy memberships derived from the virtual field reference database (VFRDB).The performance of fuzzy operators in generating fuzzy categorical maps along with the effect of land cover heterogeneity on fuzzy accuracy measures and sources of classification error were assessed. The analysis of spatial complexity computed from remote sensing images using a fractal model indicated that the various urban land cover types of the Perth metropolitan area are best represented at a resolution of 20 m (SPOT) as the fractal dimension (D) was found higher, as compared to the 25 m and 50 m resolutions of the Landsat-7 ETM+ and Landsat MSS, respectively, demonstrated the ability of the fractal model in distinguishing variations in the composition of built-up areas in the green and red bands of the satellite data, while forested areas typical of the urban fringe appear better characterised in the NIR band. Moran’s I of spatial autocorrelation was found useful in describing the spatial pattern of urban land cover types. A comparison between the D and Moran’s I of the study areas revealed a negative correlation, indicating that the higher the Moran’s I, the lesser the fractal dimension indicating a lower spatial complexity. Likewise, the results The accuracy of the fuzzy categorical maps associated with multiple spectral bands of a Landsat-7 ETM+ scene using various fuzzy operators reveals that the fuzzy gamma operator (y = 0.90) outperformed the categorical accuracy measures obtained by applying the fuzzy algebraic sum and other fuzzy operators for the City of Perth, while the accuracy measures of y value of 0.95 were found highest for the City of Melville and the City of Armadale.A comparison of the accuracy measures of the fuzzy land cover maps of the study areas indicated that the overall accuracy of the City of Perth was up to 13% higher than the overall accuracy of the City of Melville and the City of Armadale which was found 69% and 71%, respectively. The lower accuracy measures of the City of Melville and the City of Armadale was attributed to highly mixed land cover classes resulting in mixed pixels in Landsat-7 ETM+ scene. In addition, the spectral similarity among the class forest and grassland, urban and dense urban were identified as sources of classification errors. The analysis of spatial complexity using multiscale and multisource remote sensing data and the application of fuzzy set theory provided a viable methodology for assessing the appropriateness of scale selection for an urban analysis and generating fuzzy urban land cover maps from a multispectral image. It also illustrated the longstanding issue of carrying out the accuracy of the fuzzy land cover map considering the fuzzy memberships of the classified data and the reference data using a fuzzy error matrix.

This research presents models for the analysis of textural and contextual information content of multiscale remote sensing to select an appropriate scale for the correct interpretation and mapping of heterogeneous urban land cover types. Spatial complexity measures such as the fractal model and the Moran’s I index of spatial autocorrelation were applied for addressing the issue of scale, while fuzzy set theory was applied for mapping heterogeneous urban land cover types. Three local government areas (e.g. the City of Perth, the City of Melville and the City of Armadale) of the Perth metropolitan area were selected, as the dominant land covers of these areas are representative to the whole metropolitan area, for the analysis of spatial complexity and the mapping of complex land covers. Characterisation of spatial complexity of the study areas computed from SPOT, Landsat-7 ETM+, and Landsat MSS was used for assessing the appropriateness of a scale for urban analysis. Associated with this outcome, the effect of spectral resolution and land cover heterogeneity on spatial complexity, the performance of fractal measurement algorithms and the relationship between the fractal dimension and Moran’s I were identified. A fuzzy supervised approach of the fuzzy c-means algorithm was used to generate fuzzy memberships of the selected bands of a Landsat-7 ETM+ scene based on the highest spectral separability among different urban land covers (e.g. forest, grassland, urban and dense urban) as determined by a transformed divergence analysis. Fuzzy land cover maps resulting from the application of fuzzy operators (e.g. maximum, minimum, algebraic sum, algebraic product and gamma operators) were evaluated against fuzzy memberships derived from the virtual field reference database (VFRDB).
The performance of fuzzy operators in generating fuzzy categorical maps along with the effect of land cover heterogeneity on fuzzy accuracy measures and sources of classification error were assessed. The analysis of spatial complexity computed from remote sensing images using a fractal model indicated that the various urban land cover types of the Perth metropolitan area are best represented at a resolution of 20 m (SPOT) as the fractal dimension (D) was found higher, as compared to the 25 m and 50 m resolutions of the Landsat-7 ETM+ and Landsat MSS, respectively, demonstrated the ability of the fractal model in distinguishing variations in the composition of built-up areas in the green and red bands of the satellite data, while forested areas typical of the urban fringe appear better characterised in the NIR band. Moran’s I of spatial autocorrelation was found useful in describing the spatial pattern of urban land cover types. A comparison between the D and Moran’s I of the study areas revealed a negative correlation, indicating that the higher the Moran’s I, the lesser the fractal dimension indicating a lower spatial complexity. Likewise, the results The accuracy of the fuzzy categorical maps associated with multiple spectral bands of a Landsat-7 ETM+ scene using various fuzzy operators reveals that the fuzzy gamma operator (y = 0.90) outperformed the categorical accuracy measures obtained by applying the fuzzy algebraic sum and other fuzzy operators for the City of Perth, while the accuracy measures of y value of 0.95 were found highest for the City of Melville and the City of Armadale.
A comparison of the accuracy measures of the fuzzy land cover maps of the study areas indicated that the overall accuracy of the City of Perth was up to 13% higher than the overall accuracy of the City of Melville and the City of Armadale which was found 69% and 71%, respectively. The lower accuracy measures of the City of Melville and the City of Armadale was attributed to highly mixed land cover classes resulting in mixed pixels in Landsat-7 ETM+ scene. In addition, the spectral similarity among the class forest and grassland, urban and dense urban were identified as sources of classification errors. The analysis of spatial complexity using multiscale and multisource remote sensing data and the application of fuzzy set theory provided a viable methodology for assessing the appropriateness of scale selection for an urban analysis and generating fuzzy urban land cover maps from a multispectral image. It also illustrated the longstanding issue of carrying out the accuracy of the fuzzy land cover map considering the fuzzy memberships of the classified data and the reference data using a fuzzy error matrix.

Riparian systems are comprised of interacting aquatic and terrestrial elements that contribute distinctively to the natural capital of arid landscapes. Riparian vegetation is a major component of riparian systems, providing the ecosystem services required to support watershed health. The spatial and temporal distributions of riparian vegetation are influenced by hydrologic and disturbance processes operating at scales from local to regional. I believe both these processes are well suited to monitoring using synoptic and multitemporal approaches.The research in this dissertation is presented as 3 related studies. The first study focused on historical riparian dynamics related to natural disturbance and land use. Using current and historical riparian vegetation maps, we examined vegetation change within catchments of varying land use intensity. Results suggest that land use activities and wastewater subsidy affect the rate of development and diversity of riparian community typesThe second study used moderate resolution satellite imagery to monitor changes in riparian structure and pattern within a land cover change framework. We classified Landsat Thematic Mapper satellite imagery of the Upper Santa Cruz River watershed using Classification and Regression Tree (CART) models. We tested the ability of our models to capture change at landscape, floodplain, and catchment scales, centering our change detection efforts on a riparian tree die-off episode and found they can be used to describe both general landscape dynamics and disturbance-related riparian change.The third study examined historical and environmental factors contributing to spatial patterns of vegetation following two riparian tree die-offs. We used high resolution aerial imagery to map locations of individual live and dead trees and collected a suite of environmental variables and historical variables related directly and indirectly to land use and disturbance history. We tested for differences between groups of live and dead trees using Multi-response Permutation Procedures and found strong relationships between historical factors and mortality incidence.The results from these studies demonstrate the importance of examining historical information and spatial linkages across scales when monitoring riparian vegetation. From a land management perspective, the results identify the need for landscape-level, ecosystem-based management programs to maintain functioning and spatially connected riparian systems.

L'objectif de cette thèse était d'apporter une approche géophysique non linéaire à l'hydrologie urbaine. Elle a visé l'étude de la mise à l'échelle et de l'intermittence de la non-linéarité des précipitations, la réalisation d'une méthode de prévision stochastique à très court terme ("nowcast"), ainsi que son application aux processus hydrologiques dans les environnements (semi-) urbains. La partie modélisation hydrologique globale concerne la vallée de la Bièvre, zone semi-urbanisée de 110 km2 dans le sud-ouest de la région parisienne. Par conséquent, trois études différentes ont été réalisées dans cette zone à l'aide de deux modèles hydrologiques : le modèle conceptuel semi-distribué InfoWorks CS appliqué sur tout le bassin versant de Bièvre ; et le modèle physique complètement distribué Multi-Hydro, développé à l'École des Ponts ParisTech, appliqué sur deux sous-bassins versants de la Bièvre. Les principaux objectifs étaient de mieux comprendre les impacts de la variabilité spatio-temporelle des données pluviométriques en utilisant deux produits (les données radar bande-C de Météo-France avec une résolution de 1 km x 1 km x 5 min, et les données radar DPSRI band-X de l'ENPC à une résolution de 250 m x 250 m x 3.41 min) comme entrées pour les modèles, et d'identifier les capacités de chaque modèle pour traiter des données à une meilleur résolution, telles que la bande-X. Ensuite, les résultats obtenus démontrent que la fiabilité des simulations hydrologiques dépend intrinsèquement des caractéristiques des données pluviométriques. De plus, les données du radar bande-X pourraient mesurer des pics de précipitations plus élevés et le modèle complètement distribué était plus sensible à une meilleure résolution des données pluviométriques. Par la suite, des données de pluie provenant des radars météorologiques situés à des endroits complètement différents (Brésil, France, Japon) ont été analysées et comparées statistiquement afin d'améliorer la compréhension générale du comportement scalant des précipitations. De plus, le théorème d'intersection a été appliqué pour mettre en évidence les impacts de la variabilité spatiale d'un réseau virtuel de pluviomètres, qui a été généré en considérant l'emplacement des centres de masse de chaque sous-bassin versant de la vallée de la Bièvre. Ainsi, il a été possible d'identifier que la fractalité du réseau virtuel a conduit à une perte d'information importante des champs de pluie, biaisant leurs statistiques. Cela indique que le processus commun (largement retrouvé dans la littérature) de calibration des données radar à l'aide de pluviomètres devrait correctement prendre en compte cette fractalité. Enfin, une nouvelle approche de prévision stochastique immédiate a été proposée, à l'aide du modèle des multifractals universels (UM) en cascades continues. Cette méthode a été appliquée aux données des radars pluviométriques de la région amazonienne brésilienne et de Paris. Bien qu'il soit encore en développement et nécessite quelques améliorations, les premiers résultats obtenus avec ce modèle de prévision présenté ici sont vraiment encourageants et corroborent une fois de plus le besoin de données à haute résolution spatio-temporelle pour faire face aux crues soudaines
The focus of this thesis was to bring a nonlinear geophysical approach to urban hydrology. It aimed the study of rainfall non-linearity scaling and intermittency, achieving a stochastic very short-range forecast (nowcast) method, as well as its application to hydrological processes in (semi-) urban environments. The overall hydrological modelling part concerned the Bièvre Valley, which is a 110 km2 semi-urbanized area in the southwest of Paris region. Therefore, three different studies were performed within this area using two hydrological models: the conceptually-based semi-distributed model InfoWorks CS over the total Bièvre catchment, and the physically-based fully-distributed model developed at École des Ponts ParisTech called Multi-Hydro over two sub-catchments. The main goals were to better understand the impacts of spatio-temporal variability of rainfall data by using two products (the Météo-France C-band radar data with a resolution of 1 km x 1 km x 5 min; and the ENPC DPSRI X-band radar data at a 250 m x 250 m x 3.41 min resolution) as input to the models, and to identify the capacities of each model to deal with better resolution data, such as the X-band one. Then, the obtained results demonstrate that the reliability of the hydrological simulations are intrinsically dependent on rainfall data features. Moreover, the X-band radar data could measure higher peaks of rainfall rates and the fully-distributed model was more sensitive to better resolution rainfall data. Afterwards, different weather rainfall radar data from completely different sites (Brazil, France, Japan) were statistically analysed and compared in order to improve the general comprehension of rainfall scaling behaviour. In addition, the Intersection Theorem was applied to highlight the impacts of spatial variability of a virtual rain gauge network. The latter was generated by considering the location of each Bièvre Valley sub-catchment mass centre. Thus, it was possible to identify that the fractality of the virtual network led to an important information loss of the rainfall fields, biasing their statistics. This indicates that the common process (largely found in literature) of radar data calibration using rain gauges should be properly take into account this fractality. Finally, a new stochastic nowcast approach was proposed, using the continuous in scale cascade Universal Multifractals (UM) model. This method was applied to weather rainfall radar data from the Brazilian Amazon region and Paris. Although it is still under development and needs some improvements, the first results obtained with this forecast model presented here in this thesis are really encouraging and once more corroborate to the need of high spatio-temporal resolution data to cope flash floods

Las tendencias de los episodios de temperaturas extremas en las ciudades están aumentando (en frecuencia, magnitud y duración) debido al cambio climático en interacción con el efecto urbano. Otro factor relevante es que más de la mitad de la población mundial reside actualmente en áreas urbanas y se espera que esta tendencia demográfica aumente hasta un 68% a mediados de siglo. Ante esta creciente urbanización, los posibles impactos del cambio climático en las zonas urbanas se han convertido en una de las principales preocupaciones y, a su vez, un reto. Las morfologías urbanas y las propiedades térmicas de los materiales utilizados para construirlas son factores que influyen en la variabilidad climática espacial y temporal y se convierten en uno de los principales motivos de la singularidad urbana. El principal impacto en el microclima de las ciudades se caracteriza generalmente por el fenómeno de isla de calor urbano, que se refiere a que las áreas urbanas tienden a ser más cálidas que sus alrededores periurbanos y rurales, particularmente durante la noche, a consecuencia del calentamiento diferencial debido a su propia morfología y materiales. Se espera que el calentamiento global amplifique esta vulnerabilidad térmica, haciendo que los habitantes de las ciudades estén más expuestas a sufrir patologías asociadas a las elevadas temperaturas. Barcelona y su área metropolitana constituyen un buen ejemplo de megaciudad costera mediterránea (ciudades portuarias con una población superior a 1 millón de habitantes) y ya están siendo gravemente afectadas por estos efectos de acumulación de calor y al mismo tiempo de contaminación lumínica, acústica y en particular la atmosférica asociada a la actividad socioeconómica en un ecosistema con muy alta densidad de población y movilidad. El objetivo general de la tesis consiste en entender y modelizar la complejidad del comportamiento térmico urbano como herramienta de soporte a la toma de decisión. Una complejidad singular que modifica el clima provocando, en nuestras latitudes, una elevada vulnerabilidad a las altas temperaturas que se agravará debido al cambio climático global. Un segundo objetivo asociado al hecho de que se trata de un doctorado industrial, consiste en la elaboración de productos en base a las propiedades térmicas de las cubiertas urbanas y a los distintos escenarios de temperatura, transferibles como geoservício preoperacional o productivo al Institut Cartogràfic i Geològic de Catalunya. En primer lugar, se presentan diferentes metodologías para evaluar el efecto urbano y periurbano sobre la exposición a temperaturas extremas en Barcelona. Para ello se parte de la clasificación de Zonas Climáticas Locales o Local Climate Zones (LCZ) como enunciado base. Las LCZ consisten en un sistema de estandarización propuesto por Stewart y Oke (2012) para áreas urbanas y periurbanas clasificadas según sus respuestas térmicas. Cada categoría o uso del suelo es medible y comparable a partir de una combinación de parámetros geométricos, térmicos, radiativos y metabólicos que la caracterizan. En esta clasificación, las propiedades definidas de cada uso del suelo están directamente ligadas a una perspectiva térmica que permite estudiar los efectos del clima urbano con más detalle espacial y temporal. Las LCZ cartografiadas para nuestra región de interés, se introdujeron como entrada del modelo climático UrbClim, de alta resolución (100x100m), para crear series de temperaturas diarias (mediana y máxima) para el verano (1987-2016), con el fin de elaborar una cartografía de extremos asociada a las altas temperaturas. Utilizando la relación entre la mortalidad asociada a temperaturas elevadas y la distribución de la temperatura, se obtuvo la exposición al calor para cada LCZ que se combinó con datos poblacionales con el fin de crear mapas de vulnerabilidad climática. Todo ello se aplicó para el período climático observado y para finales de siglo 2071-2100 teniendo en cuenta los escenarios RCP 4.5 y RCP 8.5. En segundo lugar, y aprovechando las cartografías obtenidas de las LCZ, el trabajo aquí presentado se centró en mejorar los estándares propuestos en el ADN de las LCZ de Stewart y Oke (2012). La finalidad era la de mejorar los parámetros de entrada del modelo de dosel urbano WRF BEP+BEM y así mostrar la efectividad de los techos fríos y la vegetación en la reducción de la temperatura en el Área Metropolitana de Barcelona. Para hacerlo posible, se creó un flujo de trabajo centrado en la obtención de parámetros urbanos tales como el consumo de energía de los edificios y el calor antropogénico generado por los sistemas de aire acondicionado, la geometría de los cañones de las calles y los edificios, así como las características de la superficie (por ejemplo, albedo, capacidad calorífica, emisividad, fracción urbana y de vegetación). Los métodos se validaron para la ola de calor ocurrida en julio de 2015 durante la cual las temperaturas se mantuvieron entre 30 y 40°C durante cinco días consecutivos. Luego se simularon tres escenarios potenciales: 1) aumentando el albedo de cubiertas a 0.85 para ciertas clases urbanas, 2) aumentando el verde urbano en 255.64 ha adicionales de acuerdo con la propuesta del Plan Director Urbanístico de 2030 con dos esquemas de riego diferentes y 3) combinando las dos estrategias de mitigación (1 y 2). Los resultados, han permitido comprender muy bien el detalle del entramado urbano de nuestra región de estudio. Se han integrado en este estudio capas cartográficas de geoinformación que nos ha permitido comprender y modelizar íntegramente el comportamiento térmico, ya sea desde un enfoque climático o meteorológico. Gracias a estas mejoras en las cubiertas urbanas y sus propiedades, se ha podido caracterizar el entramado con el objetivo de poder hacer propuestas de mitigación y mejorar la resiliencia de la ciudad (disminuir su vulnerabilidad) en un entorno de cambio climático presente y futuro.
This thesis has been based on understanding and addressing the complexity of the urban framework and the important role that roofs and other anthropogenic factors play in the urban climate. A unique complexity that modifies the climate causing a greater vulnerability to high temperatures in our latitudes. However, there is still a lack of knowledge of the dynamics of all the processes and interactions observed in the urban boundary layer, as has been explained previously. As a result of these limitations, focusing mainly on the urban framework, this doctoral thesis is intended to advance this line of research. As this thesis is the result of an industrial doctorate (ID) with the Cartographic and Geological Institute of Catalonia (ICGC), its objective seeks operational applicability. In essence it envisages the development of a chain of methodological value based on the creation of workflows and programs that can be used and replicated on cartographic and remote sensing products. The aim is to improve the characterization of an urban and peri-urban region, in this case the Metropolitan area of Barcelona, by applying sustainable mitigation proposals to climate change. From an ID perspective, the first part satisfies the company or entity that finances, in this case the ICGC. The second part focuses mainly on university research based on the exercise of publishing and presenting results and methodologies obtained intended to be accepted within the international scientific community. To do this it is necessary to understand and advance in the knowledge of the Metropolitan Area of Barcelona (also applicable in other regions of the world), identifying the role of roofs and urban fractions, as well as their interrelationship with extreme temperatures. The goal is to alleviate the aforementioned effects by means of more sustainable proposals in a climate change context. Knowing that this is a subject which leads us to a great challenge at all levels, not only scientific but also political and educational.

L’alta risoluzione nel telerilevamento termico (Thermal Remote Sensing) da aereo o satellitare si rivela molto importante nell’analisi del comportamento termico delle superfici, in particolare per lo studio dei fenomeni climatici locali dello spazio urbano. La stato termico dell'ambiente urbano è oggi motivo di grande interesse per ricercatori, organi istituzionali e cittadini. Uno dei maggiori campi di studio del comportamento termico urbano interessa il problema energetico: la riduzione dei consumi e delle emissioni di CO2 è un obiettivo primario da perseguire per uno sviluppo sostenibile, spesso supportato da criteri legislativi e progetti comunitari. Su scala differente e con caratteristiche differenti, un altro degli argomenti che scuote da anni e con notevole interesse la ricerca scientifica, è il fenomeno termico urbano che prende il nome di isola di calore; questa si sviluppa non solo in conseguenza al calore sensibile rilasciato da attività antropiche, ma anche a causa della sempre maggiore conversione del territorio rurale in urbanizzato (inurbamento), con conseguente riduzione del fenomeno dell’evapotraspirazione. Oggetto di questa dissertazione è lo studio del comportamento termico delle superfici in ambito urbano, sperimentato sulla città di Bologna. Il primo capitolo si interessa dei principi e delle leggi fisiche sui quali è basato il telerilevamento effettuato nelle bende spettrali dell’infrarosso termico. Viene data una definizione di temperatura radiometrica e cinematica, tra loro legate dall’emissività. Vengono esposti i concetti di risoluzione (geometrica, radiometrica, temporale e spettrale) dell’immagine termica e viene data descrizione dei principali sensori su piattaforma spaziale per l’alta risoluzione nel TIR (ASTER e Landsat). Il secondo capitolo si apre con la definizione di LST (Land Surface Temperature), parametro del terreno misurato col telerilevamento, e ne viene descritta la dipendenza dal flusso della radiazione in atmosfera e dalle condizioni di bilancio termico della superficie investigata. Per la sua determinazione vengono proposti metodi diversi in funzione del numero di osservazioni disponibili nelle diverse bande spettrali dell’IR termico. In chiusura sono discussi i parametri che ne caratterizzano la variabilità. Il capitolo terzo entra nel dettaglio del telerilevamento termico in ambito urbano, definendo il fenomeno dell’Urban Heat Island su tutti i livelli atmosferici interessati, fornendo un quadro di operabilità con gli strumenti moderni di rilievo alle differenti scale (analisi multiscala). Un esempio concreto di studio multiscala dei fenomeni termici urbani è il progetto europeo EnergyCity, volto a ridurre i consumi energetici e le emissioni di gas serra di alcune città del centro Europa. Il capitolo quarto riporta la sperimentazione condotta sull’isola di calore urbana della città di Bologna tramite immagini ASTER con risoluzione spaziale 90 m nel TIR e ricampionate a 15 m dal VIS. Lo studio dell’isola di calore si è effettuata a partire dal calcolo della Land Surface Temperature utilizzando valori di emissività derivati da classificazione delle superfici al suolo. Per la validazione dei dati, in alternativa alle stazioni di monitoraggio fisse dell’ARPA, presenti nell’area metropolitana della città, si è sperimentato l’utilizzo di data-loggers per il rilievo di temperatura con possibilità di campionamento a 2 sec. installati su veicoli mobili, strumentati con ricevitori GPS, per la misura dei profili di temperatura atmosferica near-ground lungo transetti di attraversamento della città in direzione est-ovest.

Evapotranspiration (ET; evaporation + transpiration) is central to a wide range of biological, chemical, and physical processes in the Earth system. Accurate remote sensing of ET is challenging due to the interrelated and generally scale dependent nature of the physical factors which contribute to the process. The evaporation of water from porous media like sands and soils is an important subset of the complete ET problem. Chapter 1 presents a laboratory investigation into this question, examining the effects of grain size and composition on the evolution of drying sands. The effects of composition are found to be 2-5x greater than the effects of grain size, indicating that differences in heating caused by differences in reflectance may dominate hydrologic differences caused by grain size variation. In order to relate the results of Chapter 1 to the satellite image archive, however, the question of information loss between hyperspectral (measurements at 100s of wavelength intervals) laboratory measurements and multispectral (≤ 12 wavelength intervals) satellite images must be addressed. Chapter 2 focuses on this question as applied to substrate materials such as sediment, soil, rock, and non-photosynthetic vegetation. The results indicate that the continuum that is resolved by multispectral sensors is sufficient to resolve the gradient between sand-rich and clay-rich soils, and that this gradient is also a dominant feature in hyperspectral mixing spaces where the actual absorptions can be resolved. Multispectral measurements can be converted to biogeophysically relevant quantities using spectral mixture analysis (SMA). However, retrospective multitemporal analysis first requires cross-sensor calibration of the mixture model. Chapter 3 presents this calibration, allowing multispectral image data to be used interchangeably throughout the Landsat 4-8 archive. In addition, a theoretical explanation is advanced for the observed superior scaling properties of SMA-derived fraction images over spectral indices. The physical quantities estimated by the spectral mixture model are then compared to simultaneously imaged surface temperature, as well as to the derived parameters of ET Fraction and Moisture Availability. SMA-derived vegetation abundance is found to produce substantially more informative ET maps, and SMA-derived substrate fraction is found to yield a surprisingly strong linear relationship with surface temperature. These results provide context for agricultural applications. Chapter 5 investigates the question of mapping and monitoring rice agricultural using optical and thermal satellite image time series. Thermal image time series are found to produce more accurate maps of rice presence/absence, but optical image time series are found to produce more accurate maps of rice crop timing. Chapter 6 takes a more global approach, investigating the spatial structure of agricultural networks for a diverse set of landscapes. Surprisingly consistent scaling relations are found. These relations are assessed in the context of a network-based approach to land cover analysis, with potential implications for the scale dependence of ET estimates. In sum, this thesis present a novel approach to improving ET estimation based on a synthesis of complementary laboratory measurements, satellite image analysis, and field observations. Alone, each of these independent sources of information provides novel insights. Viewed together, these insights form the basis of a more accurate and complete geophysical understanding of the ET phenomenon.

Global demand for agricultural products is an increasingly important driver of deforestation in the Amazon Basin. This dissertation examines the consequences of agricultural expansion for stream ecosystems in the southern Amazon's agricultural frontier. At regional scales, the removal of watershed forest cover is known to change the energy balance and influence hydro-climatic cycling by altering stream flow, regional rainfall patterns, and land surface temperatures. At the landscape scale, these physical changes may be further exacerbated by land management practices that lead to the degradation of riparian forest buffers; decreases in connectivity; changes in the amount of light, nutrient, and sediment inputs; and decreases in water quality. Together, land use and management influence the quality and distribution of aquatic habitats within stream networks, potentially decreasing stream biotic integrity and resilience to further disturbances. Brazil's Mato Grosso state is one of the most actively expanding agricultural frontiers in the world and represents an ideal case study for examining the linkages among tropical deforestation, agricultural expansion, and the conservation of freshwater ecosystems. Mato Grosso accounted for 40% of deforestation in the Brazilian Amazon during the early 2000s - primarily due to the expansion of soybeans and cattle ranching. Deforestation rates have since dropped throughout the Amazon, but there is a lack of spatially explicit information about the land use transitions accompanying this decline. To address this gap, I combined government data on deforestation and production with the MODIS satellite time series to quantify the spatial-temporal dynamics of land use change in the region. Although agricultural expansion during this period slowed with declining commodity prices, the decline in deforestation is partly explained by a shift from soybean expansion into forests (26% of expansion from 2001-2005) to expansion into already cleared pasture lands (9% of expansion form 2006-2010). Beyond documenting these trends, the resulting dataset is a critical first step in evaluating the influence of land use and land use history on freshwater ecosystems at multiple scales. In the headwaters region of the Xingu River Basin, the proportion of small watersheds (microbasins) dominated by agriculture (>60% of area) increased from 20 to 40% from 2001 to 2010. At the same time, the stream network became increasingly fragmented by the removal of riparian forest buffers and installation of farm impoundments. I used high resolution satellite data (ASTER) to produce the first landscape-level documentation of farm impoundments in the region, mapping approximately 10,000 impoundments (one per 7.6 km of stream length) in 2007. At the catchment scale, I collected field data in 12 headwater streams to examine the effect of land management on instream water quality. Watershed forest cover (from MODIS), the density of impoundments (from ASTER), and the percent forest in upstream riparian buffers (from Landsat) were all associated with substantial increases in stream temperature. These increases in fragmentation and water temperature may have large cumulative effects on the stream network and reduce the ability of downstream protected areas to conserve freshwater resources. At the scale of the Amazon Basin, my analysis indicates that 30% of indigenous lands and protected areas are highly vulnerable to future reductions in hydrologic connectivity, simply because of their location within their watersheds. These impacts could be substantially mitigated through enforcement of existing legislation to protect riparian buffers and new regulations to limit the number of impoundments in emerging agricultural landscapes.