Emiliana Valentini

Spectral characterization of coastal sediments using Field Spectral Libraries, Airborne Hyperspectral Images and Topographic LiDAR Data (FHyL)

Abstract Beach dune systems are important for coastal zone ecosystems as they provide natural sea defences that dissipate wave energy. Geomorphological models of this near-shore topography require site-specific sediment composition, grain size and moisture content as inputs. Hyperspectral, field radiometry and LiDAR remote sensing can be used as tools by providing synoptic maps of these properties. However, multi-remote sensing of near-shore beach images can only be interpreted if there are adequate bio-geophysical or empirical models for information extraction. Our aim was thus to model the effects of varying sediment properties on the reflectance in both field and laboratory conditions within the FHyL (Field Spectral Libraries, Airborne Hyperspectral Images and Topographic LiDAR) procedure, using a multisource dataset (airborne Hyperspectral – MIVIS and topographic LiDAR – Hawk-eye II and field radiometry). The methodology consisted of (i) acquisition of simultaneous multi-source datasets (airborne Hyperspectral – MIVIS and topographic LiDAR – Hawk-eye) (ii) hyperspectral measurements of sediment mixtures with varying physical characteristics (moisture, grain size and minerals) in field and laboratory conditions, (iii) determination and quantification of specific absorption features, and (iv) correlation between the absorption features and physical parameters cited above. Results showed the potential of hyperspectral signals to assess the effect of moisture, grain-size and mineral composition on sediment properties.

FHYL: Field spectral libraries, airborne hyperspectral images and topographic and bathymetric LiDAR data for complex coastal mapping

Remote sensing of coastal ecosystems provides fundamental information for the effective assessment of valuable natural habitat. A synoptic view of the shallow submerged and emerged coastal landscape can offer the quantitative ability to obtain spatially explicit data over large and complex areas, where the heterogeneity of habitat is mixed by vegetation and sediment/soil interactions at the interface with water and atmosphere. In the present paper, by combining field radiometry with contemporary airborne hyperspectral imagery and topographic and bathymetric LiDAR data, an innovative approach with the application of Spectral Mixture Analysis (SMA) and Multiple Linear Regression models is proposed in order to define shallow coastal seabed, beach and dune habitat at the finest scale. The implemented FHyL (Field spectral libraries, airborne Hyperspectral imagery and LiDAR altimetry) processing chain leads to an innovative mapping results obtained by an integration of multisensory data. Presence and typology of vegetated and unvegetated beach is represented as the abundance of each physical response within the hyperspectral reflectance by building multisensory and multidimensional hyperspectral - LiDAR mixture space. Mineralogical and sedimentological proprieties of the beach sediment was estimated by using field and airborne spectral libraries combined with sediment sampling in a multiple linear regression statistical model. Therefore, FHyL represents the multisensory data fusion process to classify and map vegetation presence and distribution, as well as sediment properties and geomorphology of complex coastal seabed, and beaches dunes systems. The present research is a novel input for multilayered analysis in biophysical studies and its application on multi temporal dataset modeling of coast evolution.

Ten-years sediment dynamics in Northern Adriatic sea investigated through optical remote sensing observations

Understanding the factors influencing sediment fluxes is a key issue to interpret the evolution of coastal sedimentation under natural and human impact and relevant for the natural resources management. Despite river plumes represent one of the major gain in sedimentary budget of littoral cells, complex behavior of coastal plumes, like river discharge characteristics, wind stress and hydro-climatic variables, has not been yet fully investigated. Use of Earth Observation data allows the identification of spatial and temporal variations of suspended sediments related to river runoff, seafloor erosion, sediment transport and deposition processes. The objective of this study is to investigate superficial processes in sedimentary depositional marine environment integrating in-situ data and remote sensing data. The developed innovative approach allow quantitative evaluation of sediment dynamics using Earth Observation data, by relating spatial and temporal patterns of sediment dispersal with climatic forcings.

Detection of Natural and Anthropic Features on Small Islands

ABSTRACT Cappucci, S.; Valentini, E.; Del Monte, M.; Paci, M.; Filipponi, F., and Taramelli, A., 2017. Detection of natural and anthropic features on small islands. In: Martinez, M.L.; Taramelli, A., and Silva, R. (eds.), Coastal Resilience: Exploring the Many Challenges from Different Viewpoints. Journal of Coastal Research, Special Issue No. 77, pp. 73–87. Coconut Creek (Florida), ISSN 0749-0208. Mapping the distribution of seabed habitats, and estimating the spatial distribution of features and biocenosis over land and the seafloor, is particularly important for the analysis of human impacts. The present paper uses an innovative image analysis method that integrates different data sources from airborne remote sensing and in situ measurements for different features, allowing the detection of ecological ‘tipping points’ both in emerged and submerged coastal environments. Results show that it is possible to differentiate between the respective roles of: first, the internal variability of the natural morphological system and second, of external forcing factors. The final evidence, however, identifies a clear signature of external forcing, but whether of anthropogenic or natural origin, is unclear. The spatial pattern of the response to anthropogenic forcing may be indistinguishable from patterns of natural variability. It is argued that this novel approach to define tipping points following anthropogenic impacts could be most valuable in the management of natural resources and the economic development of coastal areas worldwide.

Modelling coastal processes by means of innovative integration of remote sensing time series analysis

Coastal marine and inland landforms are dynamic systems undergoing adjustments in form at different time and space scales in response to varying conditions external to the system. Coastal emerged and nearshore areas, affected by short-term perturbations, return to their pre-disturbance morphology and generally reach a dynamic equilibrium. The objective of this research is to propose innovative remote sensing applications for monitoring and/or combination of existing ones to monitor specific coastal processes in order to quantify and model their time evolution. In particular, it shows which properties are the best proxy for remote sensing characterisation of nearshore coastal areas both emerged and submerged environments by combining multi-sensor spaceborne remote sensing (SAR and OPTICAL) to a) produce deformation and spatiotemporal variations maps in coastal morphology with a special focus to point out the temporal subsidence evolution, b) integrate inter and intra-annual change detection maps of vegetation into coastal morphology.

A Hybrid Power Law Approach for Spatial and Temporal Pattern Analysis of Salt Marsh Evolution

ABSTRACT Taramelli, A.; Valentini, E.; Cornacchia, L., and Bozzeda, F., 2017. A hybrid power law approach for spatial and temporal pattern analysis of salt marsh evolution. In: Martinez, M.L.; Taramelli, A., and Silva, R. (eds.), Coastal Resilience: Exploring the Many Challenges from Different Viewpoints. Journal of Coastal Research, Special Issue No. 77, pp. 62–72. Coconut Creek (Florida), ISSN 0749-0208. A striking feature in salt marshes is vegetation distribution, which can self-organize in patterns over time and space. Self-organized patchiness of vegetation can often give rise to power law relationships in the frequency distribution of vegetation patch sizes. In cases where the whole distribution does not follow a power law, the variance of scale in its tail may often be disregarded. The deviation from power laws represents stochastic effect that can be hybridized on the basis of a Fuzzy Bayesian (FB) generative algorithm, to emphasize the influence of different physical and climatic variables on the patch size distribution to detect tipping point of the ephemeral life of a salt marsh under frequent disturbance events. Using remote sensing observations, we investigate the statistical distribution of spatial vegetation patterns controlled by changes in environmental variables acting on salt marshes, and we speculate the conditions under which a shift from a scale-invariant (power law) distribution to patterns characterized by a dominant patch size could be expected using channel sinuosity as a parameter. Our results show that the hybrid model without considering channel sinuosity can only detect the pure power law exponent, while considering channel sinuosity detects the exponent in the tail of the patch size distribution. Thus the evolution of vegetation patches (under power law) detected by the hybrid model considering channel sinuosity can then be used to forecast potential deviation from steady states in intertidal systems, taking into account the climatic and hydrological regimes. The research shows how numerical thresholds can describe the influence of changes in the non-linearity of patch size frequency distribution and how these thresholds can be reflected as attributes for the resilience capacity measurements in estuarine salt marshes.

Multisensory data fusion methods for the estimation of beach sediment features: Mineralogical, grain size and moisture

The research presented in this paper belongs to a wider research aimed to test innovative remote sensed techniques for the environmental and ecological characterization of emerged and submerged coastal areas [1]. Here we focus on multisensory data fusion methods for the estimation of beach sediment parameters (mineralogy, grain size and moisture content) applied to a 22 km long sandy beach, in the Sabaudia-Latina physiographic unit (central Italy) (Fig. 1). The lithologic composition and grain size distribution of sediments are primary determinants of their inherent reflectance properties [2,3]. Moreover, moisture content is also known to have a strong influence on reflectance of soils and sediments. If the effects of sediment composition, grain size and moisture content could be distinguished spectrally, it might be possible to map these properties at synoptic scales using hyperspectral, or perhaps even broadband, remote sensing in conjunction with few field sampling measures. In this study, we attempt to estimate the distribution of each of the above parameters through a multi linear regression model of airborne hyperspectral bands.

Remote Sensing Solutions to Monitor Biotic and Abiotic Dynamics in Coastal Ecosystems

Land- and seascapes are complex systems heterogeneous in space and variable in time. Land- and seascapes can be characterized by attributes and exchanges of energy and mass between these attributes. Under such conditions the most realistic interpretation of nearshore land- and seascapes is as open systems, with multiple interactions between attributes and environment.

Global MODIS Fraction of Green Vegetation Cover for Monitoring Abrupt and Gradual Vegetation Changes

The presence and distribution of green vegetation cover in the biosphere are of paramount importance in investigating cause-effect phenomena at the land/atmosphere interface, estimating primary production rates as part of global carbon and water cycle assessments and evaluating soil protection and land use change over time. The fraction of green vegetation cover (FCover) as estimated from satellite observations has already been demonstrated to be an extraordinarily useful product for understanding vegetation cover changes, for supporting ecosystem service assessments over areas with variable extents and for processes spanning a variable period of time (abrupt events or long-term processes). This study describes a methodology implemented to estimate global FCover (from 2001 to 2015) by applying a linear spectral mixture analysis with global endmembers to an entire temporal series of MODIS satellite observations and gap-filling missing FCover observations in temporal series using the DINEOF algorithm. The resulting global MODV1 FCover product was validated with two global validation datasets and showed an overall good thematic absolute accuracy (RMSE = 0.146) consistent with the validation performance of other FCover global products. Basic statistics performed on the product show changes in average and trend values and allow for the quantification of gross vegetation loss and gain over different temporal scales. To demonstrate the capacity of this global product to monitor specific dynamics, a multitemporal analysis was performed on selected sites and vegetation responses (i.e., cover changes), and specific dynamics resulting from cause-effect phenomena are briefly discussed. The product is intended to be used for monitoring vegetation dynamics, but it also has the potential to be integrated in other modeling frameworks (e.g., the carbon cycle, primary production, and soil erosion) in conjunction with other spatial datasets such as those on climate and soil type.

Figure 6-7_Patch size vs Sinuosity.xlsx

Figure 6 – Relationships between vegetation patch size (m 2 , log. scale) and maximum sinuosity of the adjacent channel, on both the high-resolution RWS vegetation map from 2010 and the two SPOT remote sensing images from 1999 and 2012. The relationships are shown for each of the three main vegetation typologies (e.g., pioneer, Elymus and reed vegetation). Figure 7 – Boxplot of the distribution of channel sinuosity values within each vegetation typology, on the RWS vegetation map and on the SPOT satellite images from 1999 and 2012. Different letters denote significant differences among the three vegetation classes (ANOVA, p = 0.05).