Refine
Has Fulltext
- yes (2) (remove)
Document Type
- Doctoral Thesis (2) (remove)
Language
- English (2) (remove)
Is part of the Bibliography
- yes (2)
Keywords
- Klassifikation (2) (remove)
The Himalayas are a region that is most dependent, but also frequently prone to hazards from changing meltwater resources. This mountain belt hosts the highest mountain peaks on earth, has the largest reserve of ice outside the polar regions, and is home to a rapidly growing population in recent decades. One source of hazard has attracted scientific research in particular in the past two decades: glacial lake outburst floods (GLOFs) occurred rarely, but mostly with fatal and catastrophic consequences for downstream communities and infrastructure. Such GLOFs can suddenly release several million cubic meters of water from naturally impounded meltwater lakes. Glacial lakes have grown in number and size by ongoing glacial mass losses in the Himalayas. Theory holds that enhanced meltwater production may increase GLOF frequency, but has never been tested so far. The key challenge to test this notion are the high altitudes of >4000 m, at which lakes occur, making field work impractical. Moreover, flood waves can attenuate rapidly in mountain channels downstream, so that many GLOFs have likely gone unnoticed in past decades. Our knowledge on GLOFs is hence likely biased towards larger, destructive cases, which challenges a detailed quantification of their frequency and their response to atmospheric warming. Robustly quantifying the magnitude and frequency of GLOFs is essential for risk assessment and management along mountain rivers, not least to implement their return periods in building design codes.
Motivated by this limited knowledge of GLOF frequency and hazard, I developed an algorithm that efficiently detects GLOFs from satellite images. In essence, this algorithm classifies land cover in 30 years (~1988–2017) of continuously recorded Landsat images over the Himalayas, and calculates likelihoods for rapidly shrinking water bodies in the stack of land cover images. I visually assessed such detected tell-tale sites for sediment fans in the river channel downstream, a second key diagnostic of GLOFs. Rigorous tests and validation with known cases from roughly 10% of the Himalayas suggested that this algorithm is robust against frequent image noise, and hence capable to identify previously unknown GLOFs. Extending the search radius to the entire Himalayan mountain range revealed some 22 newly detected GLOFs. I thus more than doubled the existing GLOF count from 16 previously known cases since 1988, and found a dominant cluster of GLOFs in the Central and Eastern Himalayas (Bhutan and Eastern Nepal), compared to the rarer affected ranges in the North. Yet, the total of 38 GLOFs showed no change in the annual frequency, so that the activity of GLOFs per unit glacial lake area has decreased in the past 30 years. I discussed possible drivers for this finding, but left a further attribution to distinct GLOF-triggering mechanisms open to future research.
This updated GLOF frequency was the key input for assessing GLOF hazard for the entire Himalayan mountain belt and several subregions. I used standard definitions in flood hydrology, describing hazard as the annual exceedance probability of a given flood peak discharge [m3 s-1] or larger at the breach location. I coupled the empirical frequency of GLOFs per region to simulations of physically plausible peak discharges from all existing ~5,000 lakes in the Himalayas. Using an extreme-value model, I could hence calculate flood return periods. I found that the contemporary 100-year GLOF discharge (the flood level that is reached or exceeded on average once in 100 years) is 20,600+2,200/–2,300 m3 s-1 for the entire Himalayas. Given the spatial and temporal distribution of historic GLOFs, contemporary GLOF hazard is highest in the Eastern Himalayas, and lower for regions with rarer GLOF abundance. I also calculated GLOF hazard for some 9,500 overdeepenings, which could expose and fill with water, if all Himalayan glaciers have melted eventually. Assuming that the current GLOF rate remains unchanged, the 100-year GLOF discharge could double (41,700+5,500/–4,700 m3 s-1), while the regional GLOF hazard may increase largest in the Karakoram.
To conclude, these three stages–from GLOF detection, to analysing their frequency and estimating regional GLOF hazard–provide a framework for modern GLOF hazard assessment. Given the rapidly growing population, infrastructure, and hydropower projects in the Himalayas, this thesis assists in quantifying the purely climate-driven contribution to hazard and risk from GLOFs.
Computer Security deals with the detection and mitigation of threats to computer networks, data, and computing hardware. This
thesis addresses the following two computer security problems: email spam campaign and malware detection.
Email spam campaigns can easily be generated using popular dissemination tools by specifying simple grammars that serve as message templates. A grammar is disseminated to nodes of a bot net, the nodes create messages by instantiating the grammar at random. Email spam campaigns can encompass huge data volumes and therefore pose a threat to the stability of the infrastructure of email service providers that have to store them. Malware -software that serves a malicious purpose- is affecting web servers, client computers via active content, and client computers through executable files. Without the help of malware detection systems it would be easy for malware creators to collect sensitive information or to infiltrate computers.
The detection of threats -such as email-spam messages, phishing messages, or malware- is an adversarial and therefore intrinsically
difficult problem. Threats vary greatly and evolve over time. The detection of threats based on manually-designed rules is therefore
difficult and requires a constant engineering effort. Machine-learning is a research area that revolves around the analysis of data and the discovery of patterns that describe aspects of the data. Discriminative learning methods extract prediction models from data that are optimized to predict a target attribute as accurately as possible. Machine-learning methods hold the promise of automatically identifying patterns that robustly and accurately detect threats. This thesis focuses on the design and analysis of discriminative learning methods for the two computer-security problems under investigation: email-campaign and malware detection.
The first part of this thesis addresses email-campaign detection. We focus on regular expressions as a syntactic framework, because regular expressions are intuitively comprehensible by security engineers and administrators, and they can be applied as a detection mechanism in an extremely efficient manner. In this setting, a prediction model is provided with exemplary messages from an email-spam campaign. The prediction model has to generate a regular expression that reveals the syntactic pattern that underlies the entire campaign, and that a security engineers finds comprehensible and feels confident enough to use the expression to blacklist further messages at the email server. We model this problem as two-stage learning problem with structured input and output spaces which can be solved using standard cutting plane methods. Therefore we develop an appropriate loss function, and derive a decoder for the resulting optimization problem.
The second part of this thesis deals with the problem of predicting whether a given JavaScript or PHP file is malicious or benign. Recent malware analysis techniques use static or dynamic features, or both. In fully dynamic analysis, the software or script is executed and observed for malicious behavior in a sandbox environment. By contrast, static analysis is based on features that can be extracted directly from the program file. In order to bypass static detection mechanisms, code obfuscation techniques are used to spread a malicious program file in many different syntactic variants. Deobfuscating the code before applying a static classifier can be subjected to mostly static code analysis and can overcome the problem of obfuscated malicious code, but on the other hand increases the computational costs of malware detection by an order of magnitude. In this thesis we present a cascaded architecture in which a classifier first performs a static analysis of the original code and -based on the outcome of this first classification step- the code may be deobfuscated and classified again. We explore several types of features including token $n$-grams, orthogonal sparse bigrams, subroutine-hashings, and syntax-tree features and study the robustness of detection methods and feature types against the evolution of malware over time. The developed tool scans very large file collections quickly and accurately.
Each model is evaluated on real-world data and compared to reference methods. Our approach of inferring regular expressions to filter emails belonging to an email spam campaigns leads to models with a high true-positive rate at a very low false-positive rate that is an order of magnitude lower than that of a commercial content-based filter. Our presented system -REx-SVMshort- is being used by a commercial email service provider and complements content-based and IP-address based filtering.
Our cascaded malware detection system is evaluated on a high-quality data set of almost 400,000 conspicuous PHP files and a collection of more than 1,00,000 JavaScript files. From our case study we can conclude that our system can quickly and accurately process large data collections at a low false-positive rate.