Maria Lanfredi

INVESTIGATING CLUSTERING STRUCTURES IN TIME-OCCURRENCE SEQUENCES OF SEISMIC EVENTS OBSERVED IN THE IRPINIA-BASILICATA REGION (SOUTHERN ITALY)

We reveal the existence of clustering properties in the temporal distribution of the earthquakes observed in a seismic active area of Southern Apennine Chain (Italy) by means of quantitatively fractal tools (Fano Factor and Allan Factor). Data consist in a sequence of seismic events instrumentally recorded during the period 1983–1995 in the Irpinia-Basilicata Region (Southern Italy), in which in past and recent years, many destructive events occurred. The analysis of the Fano and Allan Factors shows that the sequence of the occurrence times of events with magnitude Mth≥2.5 is characterized by a scale-invariant behavior from the time scale τ ~ 5 · 103 s with a scaling coefficient α ~ 0.3. By gradually increasing the threshold magnitude up to Mth=3.1, the value of the scaling coefficient monotonically decreases, pointing out a falling-off in the correlation strength. Although the increasing of the threshold magnitude seems to act as a randomizing filter which removes clustered structures, no firm sign of Poissonian, memoryless behavior is detectable in our analysis.

Early Identification of Land Degradation Hotspots in Complex Bio-Geographic Regions

The development of low-cost and relatively simple tools to identify emerging land degradation across complex regions is fundamental to plan monitoring and intervention strategies. We propose a procedure that integrates multi-spectral satellite observations and air temperature data to detect areas where the current status of local vegetation and climate shows evident departures from the mean conditions of the investigated region. Our procedure was tested in Basilicata (Italy), which is a typical bio-geographic example of vulnerable Mediterranean landscape. We grouped Landsat TM/ETM+ NDVI and air temperature (T) data by vegetation cover type to estimate the statistical distributions of the departures of NDVI and T from the respective land cover class means. The pixels characterized by contextual left tail NDVI values and right tail T values that persisted in time (2002–2006) were classified as critical to land degradation. According to our results, most of the critical areas (88.6%) corresponded to forests affected by erosion and to riparian buffers that are shaped by fragmentation, as confirmed by aerial and in-situ surveys. Our procedure enables cost-effective screenings of complex areas able to identify raising hotspots that require urgent and deeper investigations.

Landscape-scale characterization of vegetation phenology using AVHRR-NDVI and Landsat-TM data

The strict link between intra-annual vegetation dynamics (phenology) and Earths climate makes phenological information fundamental to improve understanding and models of inter-annual variability in terrestrial carbon exchange and climate-biosphere interactions. In order to monitor phenology in a landscape characterized by heterogeneous features rapidly changing over the territory, we performed multitemporal classifications of NDVI-AVHRR data and interfaced them with Landsat-TM data and orography. The sample area is the Vulture basin (Southern Italy), where cultivated and densely vegetated areas coexist with urban and recently built industrial areas. These land cover patterns rapidly change over the territory at very small spatial scales; it is a complex zone very interesting for studying the use of remote sensing techniques in the integrated monitoring context. Clusters having homogeneous NDVI time behaviors were identified. In spite of its spatial resolution, AVHRR NDVI effectively picks up the characteristic phenology for different covers and altitudes. Moreover, some pixels having particular microclimate were clustered and their characterization was only possible by using orography and TM classification information. The comparison of two intra-annual classifications (1996 and 1998) showed that the proposed approach can be very useful for studying change in pattern of vegetation dynamics.

A geostatistics-assisted approach to the deterministic approximation of climate data

We propose a nonconventional application of variogram analysis to support climate data modelling with analytical functions. This geostatistical technique is applied in the theoretical domain defined by each model variable to detect the systematic behaviours buried in the fluctuations determined by other driving factors and to verify the ability of candidate fits to remove correlations from the data. The climatic average of the atmospheric temperature measured at 387 European meteorological stations has been analysed as a function of geographical parameters by a step-wise procedure. Our final model accounts for non-linearity in latitude with a local-scale residual correlation that decays in approximately ten kilometres. The variance of the residuals from the fitted model (approximately 3% of the total) is mostly determined by local heterogeneity in transitional climates and by urban islands. Our approach is user-friendly, and the support of statistical inference makes the modelling self-consistent. Variogram analysis is adapted to optimize deterministic modelling of climate data.The explanatory variables define the domain where our analysis is performed.Variograms support model identification and diagnostic checking.The analysis identifies scales where simple functions approximate complex patterns.The approach is suited when local and global scales are separable.

Integrated Indicators for the Estimation of Vulnerability to Land Degradation

The setting up of sustainable development strategies, able to balance the opposite demands of economic growth and environmental protection, is one of the fundamental challenges for the international community. Our developing world is experiencing growing pressures on its land, water, and food production systems and the role of the human society in determin‐ ing change within the Earth environment is becoming ever more central [1]. In this context, preserving the land productivity is a prior goal, especially in those areas, such as drylands, which are particularly fragile from an ecological point of view.

Time Correlation Structure in Hourly Concentration Time Series of CO, NOx and O3 in Urban Areas

The timescales which govern urban pollution processes are investigated by analyzing variance spectra and structure functions of observational time series. The range of analyzed scales stretches from one hour to several days. It is shown that characteristic fluctuations of CO, NOx (primary pollutants) and O3 (secondary pollutant) follow a scale invariant law up to timescales of about one day. Scaling exponents indicate the presence of stabilizing feedback mechanisms. Such a scale invariance is broken by the appearance of basic periods which, for primary pollutants, are expressions of traffic dynamics, whereas, for ozone, are closely linked to the diurnal and annual solar cycles.

Time Correlation Laws Inferred from Climatic Records: Long-Range Persistence and Alternative Paradigms

Observational time series of climatic variables exhibit substantial changeability on spatial and temporal scales over many orders of magnitude. In statistical terms, this implies a continuous variance distribution involving all resolvable time scales (frequencies), starting from those comparable with the age of the Earth. A correct causal interpretation of such a variability is very difficult even in the context of a cognitive approach (e.g., von Storch, 2001) to the problem. Cognitive models are minimum complexity models aiming at the scientific understanding of the most relevant processes occurring at any given temporal and spatial scale. Although generally they cannot be useful for management decisions straightforwardly, their role is fundamental especially for understanding the internal climatic variability that cannot be passively related to external forcing factors. The concept of stochastic process is essential in this framework, since it synthesizes collective behaviours which contribute as a whole to the overall dynamics. As stochastic processes are the macroscopic result of many degrees of freedom, the characterization of their correlation properties across different scales through the analysis of observational data is a problem of statistical inference and their modelling is usually a mechanical-statistical problem. Maybe, the most famous early effort aiming to summarize the climate variance distribution among different frequencies, which is commonly referred as climate spectrum, is the ideal sketch proposed by Mitchell (1976) (see Fig. 1). All the features of this spectrum that deviate from the flat behaviour typical of white noise (pure random process) deserve dynamical interpretation in order to understand climate. Within the traditional picture of the climate dynamics, the variance distribution among different temporal scales is seen as the superposition of oscillations generated by astronomical cycles (spectral spikes), quasi-periodic or aperiodic fluctuations with a preferred scale (broad spectral peaks), and internal stochastic processes whose temporal correlation decays according to characteristic time scales. These last are responsible for all the continuous broad-band deviations of the spectrum from flatness. Within this picture, the variance accumulations that do not appear in the form of peaks and spikes, such as that we

SEARCHING FOR PERSISTENCE IN ATMOSPHERIC TEMPERATURE TIME SERIES: A RE-VISITATION OF RESULTS FROM DETRENDED FLUCTUATION ANALYSIS

We re-analyze historical daily atmospheric temperature time series for investigating long-range correlation. Such a problem is attracting much attention due to the deep importance of assessing statistical dependence of atmospheric phenomena on climatic scales in the context of Climate modeling. In particular, we adopt Detrended Fluctuation Analysis (DFA), which is one of the most used techniques for detecting scale invariance in stationary signals contaminated by external non-stationary disturbances. A very standard application of this methodology seems to evidence persistence power-law exponents close to 0.65 on time scales greater than the meteorological one (>15 days). Nevertheless, more careful investigations put into evidence the local character of this exponent whose value decays progressively with scale. Our results show that the simple detection of approximately straight lines in a log–log plot cannot be considered as a signature of scale invariance and local scale features have to be explicitly investigated.

Comparison of MIVIS and IKONOS data for high-resolution land-cover classification in a rural/mountainous area

Land cover classification is one of the main applications of remotely sensed data and the capability of airborne hyperspectral data for such a purpose is known. The recent availability of high spatial resolution multispectral data, such as IKONOS and QuickBird, puts the question about advantages and disadvantages of these data in comparison with the hyperspectral ones. We evaluated the cost and accuracy of using IKONOS imagery to perform a land cover classification at high spatial resolution and compared them with results obtained from MIVIS airborne hyper-spectral scanner data (102 bands from VIS to TIR). The study was performed in a rural area (25 km2) of Basilicata region (Southern Italy) characterized by complex topography (altitude ranges from 600 to 1400m) and different land cover patterns (forests, lakes, cultivated areas, and small urban areas). Evaluations were made taking into account time-processing, feature extraction, accuracy for different classification levels, and costs as a function of the extension of the area to be classified. Quite high accuracies were obtained for the first classification level, whereas increasing the class number IKONOS was less accurate than MIVIS. Multispectral classification well identified the different forest species, but had some problems in distinguishing between gravel road and some plowed lands. The obtained results showed that IKONOS data are cost-effective for updating thematic maps to support planning and decision-making processes at local government scale.

A Digression on the Analysis of Historical Series of Daily Data for the Characterization of Precipitation Dynamics

Precipitation, together with temperature, is the most important variable in defining the climate of a region. Then, the right understanding of rainfall variability, which occurs over a wide range of temporal scales, has relevance for a large variety of problems linked to meteorology and climate, both in theoretical and practical frameworks. The double aspect, continuous and point process, of rainfall sequences manifests itself depending on the scale of aggregation of the rainfall events and on the intensity thresholds associated to storminess risk. This requires the use of different characteristic variables, different reference models as well as different analysis techniques for obtaining a comprehensive characterization of the observational time series and assessing risk. This Chapter provides a quick overview of the many aspects of the reconstruction of the time scale properties based on the investigation of historical data. Storminess observed for several decades at two Italian sites (Genoa and Palermo), which exhibit different climatic features, were analysed both with tools typical of point processes and more standard analysis techniques to provide a coherent picture of the basic properties of rainfalls that can be extracted from daily data about weather, seasonal, and climatic scales. Both analogous and complementary cycles appear when we approach the problem from the two different perspectives separately; additional behaviours are detected when we integrate them. This comprehensive picture of historical data represents the background for understanding precipitation regimes and identifying possible climatic changes or human pressure effects that could increase storminess risk.