Alexander Kolovos

Total ozone mapping by integrating databases from remote sensing instruments and empirical models

Atmospheric studies often require the generation of high-resolution maps of ozone distribution across space and time. The high natural variability of ozone concentrations and the different levels of accuracy of the algorithms used to generate data from remote sensing instruments introduce major sources of uncertainty in ozone modeling and mapping. These aspects of atmospheric ozone distribution cannot be confronted satisfactorily by means of conventional interpolation and statistical data analysis. We suggest that the techniques of Modern Spatiotemporal Geostatistics (MSG) can be used efficiently to integrate salient (although of varying uncertainty) physical knowledge bases about atmospheric ozone in order to generate and update realistic pictures of ozone distribution across space and time. The MSG techniques rely on a powerful scientific methodology that does not make the restrictive modeling assumptions of previous techniques. A numerical study is discussed involving datasets generated by measuring instruments onboard the Nimbus 7 satellite. In addition to exact (hard) ozone data, the MSG techniques process uncertain measurements and secondary (soft) information in terms of total ozone-tropopause pressure empirical relationships. Nonlinear estimators are used, in general, and non-Gaussian probability laws are automatically incorporated. The proposed total ozone analysis can take into consideration major sources of error in the Total Ozone Mapping Spectrometer solar backscatter ultraviolet tropospheric ozone residual (related to data sampling, etc.) and produce high spatial resolution maps that are more accurate and informative than those obtained by conventional interpolation techniques.

Multi-perspective analysis and spatiotemporal mapping of air pollution monitoring data.

Space-time data analysis and assimilation techniques in atmospheric sciences typically consider input from monitoring measurements. The input is often processed in a manner that acknowledges characteristics of the measurements (e.g., underlying patterns, fluctuation features) under conditions of uncertainty; it also leads to the derivation of secondary information that serves study-oriented goals, and provides input to space-time prediction techniques. We present a novel approach that blends a rigorous space-time prediction model (Bayesian maximum entropy, BME) with a cognitively informed visualization of high-dimensional data (spatialization). The combined BME and spatialization approach (BME-S) is used to study monthly averaged NO2 and mean annual SO4 measurements in California over the 15-year period 1988-2002. Using the original scattered measurements of these two pollutants BME generates spatiotemporal predictions on a regular grid across the state. Subsequently, the prediction network undergoes the spatialization transformation into a lower-dimensional geometric representation, aimed at revealing patterns and relationships that exist within the input data. The proposed BME-S provides a powerful spatiotemporal framework to study a variety of air pollution data sources.

A study of the spatiotemporal health impacts of ozone exposure

Exposure analysis and mapping of spatiotemporal pollutants in relation to their health effects are important challenges facing environmental health scientists and integrated assessment modellers. In this work, a methodological framework is discussed to study the impact of spatiotemporal ozone (O3) exposure distributions on the health of human populations. The framework, however, is very general and can be used to study various other pollutants. The spatiotemporal analysis starts with exposure distributions producing the input to pollutokinetic (or toxicokinetic) laws which are linked to effect models which, in turn, are integrated with relationships that describe how effects are distributed across populations. Important characteristics of the environmental health framework are holisticity and stochasticity. Holisticity emphasizes the functional relationships between composite space/time O3 maps, pollutokinetic models of burden on target organs and tissues, and health effects. These relationships offer a meaningful physical interpretation of the exposure and biological processes that affect human exposure. Stochasticity involves the rigorous representation of natural uncertainties and biological variations in terms of spatiotemporal random fields. The stochastic perspective introduces a deeper epistemological understanding in the development of improved models of spatiotemporal human exposure analysis and mapping. Also, it explicitly determines the knowledge bases available and develops logically plausible rules and standards for data processing and human exposure map construction. The proposed approach allows the horizontal integration among sciences related to the human exposure problem that leads to accurate and informative spatiotemporal maps of O3 exposure and effect distributions and an integrative analysis of the whole risk case. By processing a variety of knowledge bases, the spatiotemporal analysis can bring together several sciences which are all relevant to the aspect of human exposure reality that is examined.

Spatiotemporal Infectious Disease Modeling: A BME-SIR Approach

This paper is concerned with the modeling of infectious disease spread in a composite space-time domain under conditions of uncertainty. We focus on stochastic modeling that accounts for basic mechanisms of disease distribution and multi-sourced in situ uncertainties. Starting from the general formulation of population migration dynamics and the specification of transmission and recovery rates, the model studies the functional formulation of the evolution of the fractions of susceptible-infected-recovered individuals. The suggested approach is capable of: a) modeling population dynamics within and across localities, b) integrating the disease representation (i.e. susceptible-infected-recovered individuals) with observation time series at different geographical locations and other sources of information (e.g. hard and soft data, empirical relationships, secondary information), and c) generating predictions of disease spread and associated parameters in real time, while considering model and observation uncertainties. Key aspects of the proposed approach are illustrated by means of simulations (i.e. synthetic studies), and a real-world application using hand-foot-mouth disease (HFMD) data from China.

Bayesian maximum entropy approach and its applications: a review

The present paper reviews the conceptual framework and development of the Bayesian Maximum Entropy (BME) approach. BME has been considered as a significant breakthrough and contribution to applied stochastics by introducing an improved, knowledge-based modeling framework for spatial and spatiotemporal information. In this work, one objective is the overview of distinct BME features. By offering a foundation free of restrictive assumptions that limit comparable techniques, an ability to integrate a variety of prior knowledge bases, and rigorous accounting for both exact and uncertain data, the BME approach was coined as introducing modern spatiotemporal geostatistics. A second objective is to illustrate BME applications and adoption within numerous different scientific disciplines. We summarize examples and real-world studies that encompass the perspective of science of the total environment, including atmosphere, lithosphere, hydrosphere, and ecosphere, while also noting applications that extend beyond these fields. The broad-ranging application track suggests BME as an established, valuable tool for predictive spatial and space–time analysis and mapping. This review concludes with the present status of BME, and tentative paths for future methodological research, enhancements, and extensions.

A GIS tool for spatiotemporal modeling under a knowledge synthesis framework

In recent years, there has been a fast growing interest in the space–time data processing capacity of Geographic Information Systems (GIS). In this paper we present a new GIS-based tool for advanced geostatistical analysis of space–time data; it combines stochastic analysis, prediction, and GIS visualization technology. The proposed toolbox is based on the Bayesian Maximum Entropy theory that formulates its approach under a mature knowledge synthesis framework. We exhibit the toolbox features and use it for particulate matter spatiotemporal mapping in Taipei, in a proof-of-concept study where the serious preferential sampling issue is present. The proposed toolbox enables tight coupling of advanced spatiotemporal analysis functions with a GIS environment, i.e. QGIS. As a result, our contribution leads to a more seamless interaction between spatiotemporal analysis tools and GIS built-in functions; and utterly enhances the functionality of GIS software as a comprehensive knowledge processing and dissemination platform.

Stochastic flowpath analysis of multiphase flow in random porous media

Multiphase flow in random porous media is studied by means of a stochastic flowpath approach, which is built upon concepts of differential geometry. This approach replaces the partial differential equations of flow by a set of ordinary differential equations along the flowpaths. Geo- metrical characterization of multiphase flow involves space transformations. The stochastic flowpath approach accounts for heterogeneity and allows for random initial or boundary conditions. It does not require perturbative approximations, and it does not involve Greens functions. The differential geometric formulation involves tracking the flowpath trajectories of the different phases within the flow domain. We discuss the solution of the two-phase flow equation within a random permeability medium by numerical simulation.

Advanced space-time predictive analysis with STAR-BME

Stochastic analysis and prediction is an important component of space-time data processing for a broad spectrum of Geographic Information Systems scientists and end users. For this task, a variety of numerical tools are available that are based on established statistical techniques. We present an original software tool that implements stochastic data analysis and prediction based on the Bayesian Maximum Entropy methodology, which has attractive advanced analytical features and has been known to address shortcomings of common mainstream techniques. The proposed tool contains a library of Bayesian Maximum Entropy analytical functions, and is available in the form of a plugin for the Quantum GIS open source Geographic Information System software.

High Resolution Ozone Mapping Using Instruments on the Nimbus 7 Satellite and Secondary Information

The high natural variability of ozone concentrations across space-time and the different levels of accuracy of the algorithms used to generate data from measuring instruments can not be confronted satisfactorily by conventional interpolation techniques. This work suggests that the Bayesian Maximum Entropy (BME) method can be used efficiently to assimilate salient (although of varying uncertainty) physical knowledge bases about atmospheric ozone in order to generate and update realistic pictures of ozone distribution. On theoretical grounds, BME relies on a powerful scientific methodology that does not make any of the restrictive modelling assumptions of previous techniques and integrates a wide range of knowledge bases. A study is discussed in which BME assimilates data sets generated by measuring instruments on board the Nimbus 7 satellite as well as uncertain measurements and secondary information in terms of total ozonetropopause pressure empirical equations. The BME total ozone analysis eliminates major sources of error and produces high spatial resolution maps that are more accurate and informative than those obtained by conventional interpolation techniques.

Generating high spatial resolution analyses of SBUV stratospheric ozone for calculating the tropospheric ozone residual (TOR)

In the last decade, daily analyses of the Tropospheric Ozone Residual (TOR), which is an estimate of the vertically- integrated ozone in the troposphere, has been calculated as the difference between the vertically-integrated stratospheric ozone using data from the Solar Backscatter Ultraviolet (SBUV) remote sensing system and the total ozone from the Total Ozone Mapping Spectrometer (TOMS). Comparison of daily values of the TOMS/SBUV TOR with daily values of the surface ozone concentration and of the vertically-integrated ozone in the troposphere using ozonesonde data provided poor correlations. Reasonably good correlations were noted for longer-term (monthly, seasonally, and annually) averaged data. One of the major problems in applying SBUV data with TOMS data to develop daily estimates of the TOR is the difference in the spatial resolution. The SBUV instrument is a non-scanning, downward- looking radiometer. Data are only collected with 200-km spatial resolution along the orbital track of the satellite on which the instrument resides. The orbital tracks are as much as 25˚ longitude apart. The TOMS total ozone data, on the other hand, are collected globally on a daily basis at 50 km spatial resolution. The SBUV data gaps have been traditionally filled using conventional interpolation procedures so that the stratospheric ozone from the SBUV instrument would be available at the data locations of the TOMS instrument. Conventional interpolation procedures that have been used to fill the SBUV data gaps (e.g., linear and higher order spatial regression, kriging, basis functions, neural networks) have lacked the scientific methodology to include rigorously essential sources of physical knowledge and the conceptual organization to account for composite space-time variability effects; and, therefore, lack the ability to account for features that may exist between SBUV data sampling tracks. This factor is a cause of major errors found in the daily values of the TOMS/SBUV TOR. The objective of this study is to find an interpolation procedure that will provide significantly improved analyses of SBUV stratospheric ozone in the regions defined by the SBUV data gaps than is presently be acquired using conventional interpolations procedures. For this study, the Bayesian Maximum Entropy (BME) interpolation procedure of Modern Spatiotemporal Geostatistics was used to integrate efficiently salient physical knowledge about ozone in order to generate realistic analyses of ozone distribution across space and time. In addition to the satellite ozone measurements, BME interpolation procedure used secondary (soft) information such as the total ozone-tropopause pressure empirical relationship. The results suggested that BME interpolation procedure could eliminate a major source of error in the TOMS/SBUV TOR analyses (i.e., interpolation error), producing high spatial resolution analyses that are more accurate and informative than those presently produced using conventional interpolation techniques.