Master of Science / Department of Electrical and Computer Engineering / Caterina Maria Scoglio / Japanese Encephalitis (JE) is a vector-borne disease transmitted by mosquitoes and maintained in birds and pigs. An interconnected network model is proposed to examine the possible epidemiology of JE in the USA. Proposed JE model is an individual-level network model that explicitly considers the feral pig population and implicitly considers mosquitoes and birds in specific areas of Florida, North Carolina, and South Carolina. The virus transmission among feral pigs within a small geographic area (<60 sq mi areas) are modeled using two network topologies— fully connected and Erdos-Renyi networks. Connections between locations situated in different states (interstate links) are created with limited probability and based on fall and spring bird migration patterns. Simulation results obtained from the network models support the use of the Erdos-Renyi network because maximum incidence occurs during the fall migration period which is similar to the peak incidence of the closely related West Nile virus (WNV), another virus in the Japanese Encephalitis group (Flaviviridae) that is transmitted by both birds and mosquitoes. Simulation analysis suggested two important mitigation strategies: for low mosquito vectorial capacity, insecticidal spraying of infected areas reduces transmission and limits the outbreak to a single geographic area. Alternatively, in high mosquito vectorial capacity areas, birds rather than mosquitoes need to be removed/controlled.

The purpose of this study was to determine whether defoliation, a condition of stress caused by Choristoneura fumiferana herbivory, induces a disturbed pathologic condition in presumably normal tissue produced by the affected balsam fir trees. Techniques used in this investigation included dry weight determination, light microscopy, scanning electron microscopy (SEM), transmission electron microscopy (TEM), and biochemical and physical analyses. Terminal branch sections that contained shoots with buds and needles were cut from the upper crown on opposite sides of the tree. Collections were made from several previously selected trees in Vermont during April, June and July. The collected materials were labelled, placed in plastic bags and returned to the laboratory for histological and electron microscopic studies. The shoots derived from buds of stressed trees were shorter and produced fewer needles than shoots derived from equivalent position in nondefoliated buds. Unlike July-collected buds, the April buds did not exhibit distinct apical meristem zonation. The ultrastructural study of defoliated April-collected buds differed from nondefoliated buds, in that nondefoliated buds, were very active prior to bud burst and shoot elongation. This activity was indicated by the substantial increase in the number of Golgi-associated vesicles, several large vacuoles with heavy electron dense granular inclusions, abundant rough endoplasmic reticulum, proplastids and cell plate formation. Defoliated bud cells, unlike the nondefoliated bud cells did not contain a high population of cells exhibiting Golgi-associated rich regions, mitochondria, and apparent ce 11 p 1 ate formation. The absence or reduced presence of these metabolic indicators, in defoliated Aprilcollected buds suggest that the defoliated buds are metabolically disturbed. The eventual elongation of shoots of reduced length and number of needles as observed in June-collected buds indicates that activity was not completely inhibited in defoliated buds. The smaller bud size in defoliated balsam fir could also mean that prior accumulation of substances by the bud was effectua1 in that bud development may have been delayed. These findings indicate that Choristoneura fumiferana herbivory alters bud development.

Two important trends in computer systems are that applications are moved to the Internet as web applications, and that computer systems are getting an increasing number of cores to increase the performance. It has been shown that JavaScript in web applications has a large potential for parallel execution despite the fact that JavaScript is a sequential language. In this thesis, we show that JavaScript execution in web applications and in benchmarks are fundamentally different and that an effect of this is that Just-in-time compilation does often not improve the execution time, but rather increases the execution time for JavaScript in web applications. Since there is a significant potential for parallel computation in JavaScript for web applications, we show that Thread-Level Speculation can be used to take advantage of this in a manner completely transparent to the programmer. The Thread-Level Speculation technique is very suitable for improving the performance of JavaScript execution in web applications; however we observe that the memory overhead can be substantial. Therefore, we propose several techniques for adaptive speculation as well as for memory reduction. In the last part of this thesis we show that Just-in-time compilation and Thread-Level Speculation are complementary techniques. The execution characteristics of JavaScript in web applications are very suitable for combining Just-in-time compilation and Thread-Level Speculation. Finally, we show that Thread-Level Speculation and Just-in-time compilation can be combined to reduce power usage on embedded devices.

Syftet med den här studien är att studera vindprofiler och klassificera Low Level Jets (LLJ) och även analysera den termiska vinden över Östergarnsholm. Östergarnsholm är en liten ö som ligger 4 kilometer öster om Gotland. Viseringar och sonderingar från åren 1995 till 2001 samt år 2003 har använts. Det kriteriet som har använts för att klassificera LLJs är att det ska finnas ett vindhastighetsmaximum under 500 meters höjd. De undersökta orsakerna till uppkomsten av en LLJ är sjöbris, tröghetssvängning och termisk vind. För att kunna ta ut LLJs orsakade av sjöbris har vindprofiler över hela dygn använts. Vindens vridning vid marken, under dagen, jämfört med den överlagrade vindens riktning vid 2000 meters höjd har undersökts. LLJs som uppkommit på grund av tröghetssvängning har analyserats. Metoden för tröghetssvängning bygger på att analysera den geostrofiska vindens hastighet och vindriktning för att sedan kunna räkna ut hur vinden har blåst och var en eventuell LLJ skulle kunna bildas. Här har vindriktningar mellan 20º – 220º använts eftersom det medför att vinden då blåser från havet. Resterande vindriktningar medför att vinden har blåst över Gotland vilket resulterar i att en eventuell tröghetssvängning skulle störas och en LLJ skulle försvinna innan den når Östergarnsholm. Vindhastigeter och vindriktningar har jämförts med teoretiskt uträknade värden från tryckmätningar. Från alla 245 viseringar fanns 103 vindprofiler med LLJs. Utav dessa var 27 stycken, under 12 dygn, orsakade av sjöbris. Hur många LLJs som bildats av tröghetssvängning är oklart. Detta eftersom olika resultat erhålles beroende på om vindhastigheter och vindriktningar tas från viseringarna eller är beräknade från tryckmätningar, samt om beräkningarna av transporttiden sker med raka eller krökta trajektorier. Totalt hittades 9 LLJs som orsakats av tröghetssvängning. Metoden som använts för att analysera tröghetssvängning ger förmodligen bättre resultat över land än över hav. Detta eftersom det är svårt att mäta de exakta vindförhållandena längs luftens transport över hav. Den geostrofiska vindens ändring med höjden över ön, det vill säga den termiska vinden, har analyserats genom att undersöka alla vindprofiler, även de utan LLJs. Dessa har jämförts med den geostrofiska vinden beräknad från tryckmätningar, som representerar vinden vid marken, för att se om det uppkommer termiska vindar över Östergarnsholm. En LLJ som orsakats av termisk vind uppkommer när den geostrofiska vinden avtar med höjden, det vill säga vid negativ termisk vind. Antalet fall där den geostrofiska vinden avtar med höjden och där vinden är konstant med höjden var ungefär lika många. Däremot fanns något fler fall där den geostrofiska vinden ökade med höjden, dessa uppgick även till högre hastigheter än när vinden avtog med höjden. Det finns inget samband för vindvridningen med höjden då det förekommer LLJs på grund av termisk vind. Däremot finns ett tydligt samband mellan den negativa termiska vindens u- och v-komponent och den geostrofiska vindens u- respektive v-komponent. I båda fallen så tenderar vinden att gå mot noll med höjden. Totalt hittades 41 LLJs som var orsakade av termisk vind. Vid ungefär 50% av alla vindprofiler, både när den geostrofiska vinden avtog och ökade med höjden, var den termiska vindens nord-sydliga komposant positiv och den ost-västliga komposanten negativ. Detta ger att den varmare luften finns i nordost. / The aim of this study is to classify Low Level Jets (LLJ) and analyze the thermal wind over Östergarnsholm. Östergarnsholm is a small island that is situated 4 kilometres east of Gotland in the Baltic Sea. Pibal trackings and soundings from 1995 to 2001 and 2003 have been used in the study. The criteria that have been used to classify the LLJs is that there must be a maximum of the wind speed below 500 meters. Wind profiles from a specific day have been used to determine if the LLJs is caused by sea breeze. The shift of wind direction at ground level, during the day, compared to the geostrophic wind at 2000 meters has been analyzed. LLJs caused by internal oscillation have been analyzed. In the used method the geostrophic wind speed and wind direction have been analyzed to determine how the wind has blown and where a LLJ possibly could be formed. Only wind directions between 20º and 220º have been used to avoid that the wind should have passed Gotland. If the winds have passed Gotland the internal oscillation has been disturbed and no LLJ can appear. Measured wind speeds and directions have been compared to theoretical values from pressure measurements. 103 wind profiles with LLJs were found from all pibal trackings. In 27 of these pibal trackings, from 12 days, were LLJs caused by sea breeze. It is difficult to say how many LLJs that are caused by an internal oscillation. This due to the different results obtained when using wind speeds and wind directions from pibal trackings or pressure measurements. Also using straight line trajectories or trajectories following the isobars gives different results. Totally 9 LLJs were found, caused by an internal oscillation. Probably this way of analyze the internal oscillations is a better method for measurements over land than over sea. This due to the difficulties in measuring the exact wind speed and wind direction over the whole traveling distance over sea. The thermal wind over the island has been analyzed by studying all pibal trackings, even those without LLJs. The pibal trackings have been compared to the geostrophic wind calculated from pressure measurements to determine if thermal winds occur over Östergarnsholm. A LLJ caused by thermal wind occur when the geostrophic wind decreese with height, i.e. negative thermal wind. There was no difference in the amount of negative thermal winds and cases with no thermal wind. There were a few more positive thermal winds and they had higher wind speeds than the negative ones. There is no connection in the shift of wind direction when there is a LLJ caused by thermal wind. But there is a significant connection between the u- and v-component of the negative thermal winds and the u- and v-component of the geostrophic wind. In both cases the geostrophic wind tends to decreese towards zero with height. 41 of all LLJs were caused by thermal winds. About 50% of both negative and positive thermal winds have a positive v-component and a negative u-component. This means that the warmer region is in the north east.

Diabetes is a world-wide health problem with 415 millions of people suffering from the disease. Most diabetics are suffering from Type 2 Diabetes, which is preceded by insulin resistance in glucose utilizing tissues, such as adipose, liver, and muscle tissues. Diabetes is diagnosed when the insulin control of the glucose levels fails, which leads to high glucose levels in the blood. To better understand the insulin control of blood glucose, mathematical modeling has been used for many years to simulate the dynamics of glucose and insulin levels in the blood. Models have also been used to understand the intracellular insulin-signaling network in the insulin responding tissues. There have also been attempts to connect models from these different layers of control into a multi-level and multi-scale simulation model. However, to do such connections, several assumptions must be made about the comparability of the data from the different levels. Here, I aim for a deeper understanding of these assumptions and to use more advanced data for glucose uptake dynamics than in earlier work. I used data from the literature for the dynamics of glucose uptake in adipose and muscle tissues and improve the model in several steps to have a better agreement with these data. In particular, I refined the sub-division of the glucose uptake between the organs, to also account for liver uptake, a correction that implied a reduction by 50% for the muscle and adipose tissue glucose uptake. Unlike previous models, the updated model also describes blood flow. Finally, because of the connection to the intracellular level, the model can be used to simulate the response to anti-diabetic drugs.

The present study was designed to investigate the effects of two noise conditions, continuous and noncontinuous noise, and three intensity levels, upon a serial learning task and the interaction effects of these factors on learning.

The problem of this study was to determine the relationship between selected factors and the grade-level preference of elementary student teachers. The factors studied were self-concept, academic achievement, commitment to teaching, and personality traits.

Vid luftens passage från en typ av yta till en annan byggs ett internt gränsskikt upp där luftens egenskaper förändras successivt. Detta är grunden för bl.a. bildandet av advekterade low-level jets orsakade av en tröghetssvängning, vilken har sitt ursprung i att under vissa förhållanden erhålls obalans mellan de krafter som påverkar vinden. Målet med arbetet var att kartlägga inflytandet av low-level jets på vindprofilen på Näsudden, Gotland, ovan 100 m höjd. För detta syfte användes mätningar under perioden maj 2000 till juli 2001 gjorda i tre mätmaster i området. En mätmast var placerad på land 1.5 km från Näsuddens kust-linje, en andra mast precis vid kustlinjen och en tredje mast ute i Östersjön 4.6 km från kusten. Eftersom masten ute i havet var placerad mitt i Bockstigens vindpark, omgiven av fem vindturbiner, påverkades vindmätningarna av turbinvakar i vissa riktningssektorer. Genom jämförelser mellan vindmätningar i masten och på turbinerna i för masten ostörda sektorer, togs en korrektion fram så att vind-mätningarna gjorda på turbinerna kunde användas i de sektorer mastmätningarna stördes av turbinvakar. För att kunna studera påverkan av low-level jets på vindprofilen användes data från Bockstigens vindpark tillsammans med en vindmodell, byggd på rörelseekvationerna, Monin-Obukhovs similaritetsteori och K-teori. Modellen, som ej kan generera low-level jets, användes för att beräkna vindprofilerna över havet och hur dessa vind-profiler modifierades vid passagen av kusten. Resultatet från modellen jämfördes sedan med observationer från den 145 m höga masten på Näsudden, 1.5 km från kusten. Det visade sig att de observerade vindhastigheterna på 145 m höjd i medeltal var ca 0.8 m/s högre än vad som beräknats med modellen vid vindriktningar från havet in över Näsudden, vilket kan antas bero på inflytandet från low-level jets. Under månaderna mars till maj är vattnet i Ösersjön klimatologiskt sett mycket kallare än luften, ett förhållande som är fördelaktigt för bildandet av low-level jets. Det visade sig att under dessa månader var vindhastigheten mätt på 145 m höjd inom vind-riktningssektorn 220°-250° hela 1.7 m/s högre än vad som beräknades med vind-modellen, vilket understryker betydelsen av low-level jets under denna del av året. Under resten av året befanns skillnaden mellan observerad och modellberäknad vindhastighet på 145 m höjd vara ca 0.8 m/s för samma vindriktningar. / When air is flowing from one type of surface to another, an internal boundary layer is built up where the properties of the air gradually change. This is the basic condition for the development of for example advected low-level jets caused by inertial oscillations, whose origin is that under certain conditions an imbalance will occur between the forces driving the wind. The purpose with the present work was to map the influence from low-level jets on the wind profile at Näsudden, Gotland, above 100 m height. For this purpose, measurements taken on three towers in the area during the period May 2000 to July 2001 were used. One tower was located inland 1.5 km from the coastline of Näsudden, a second tower was located just at the coastline, and a third one was located offshore 4.6 km from the coast. Since the offshore tower was located in the middle of the Bockstigen wind farm, surrounded by five wind turbines, the wind measurements were affected by turbine wakes in certain wind direction sectors. By comparing the wind measurements taken on the tower with measurements taken on the turbines for directions where the tower was not in the wake, corrections were determined where by the wind measurements taken on the turbines could be used for sectors where the tower was located in wakes. To investigate the influence from low-level jets on the wind profile, data from Bockstigen were used together with a wind model based on the equations of motion, Monin-Obukhov’s similarity theory, and K-theory. The model, which is unable to generate low-level jets, was used to calculate the wind profiles over the sea and how these profiles were modified after passage of the coastline. The results from the model were compared to observations made on the 145 m high tower at Näsudden, 1.5 km from the coast. The comparison showed that the observed wind speed at 145 m height on the average was ca 0.8 m/s higher than the wind speed calculated with the model for wind directions from the sea at Näsudden, which could be assumed to be the effect of the presence of low-level jets. During the months March to May the water in the Baltic Sea is climatologically colder than the air temperature, which is favorable for the development of low-level jets. The results from these months showed that the measured wind speed at 145 m height in the wind direction sector 220°-250° was as much as 1.7 m/s higher than what was calculated with the model, which points at the importance of low-level jets during this part of the year. During the rest of the year the average difference between observed and modeled wind speed at the 145 m level was about 0.8 m/s for the same wind direction.

This research study examines two recent conceptualisations of leadership that have evolved from previous leadership research. They are the Level 5 leadership theory and the romance of leadership theory. These concepts have been chosen as they represent conflicting perspectives of leadership. The aim of this research study is to investigate whether a recent historical profile of publicity given to a prominent Level 5 leader lends itself to a romance of leadership interpretation. The present study addressed this aim by adopting a research methodology that is qualitative in nature and by utilising archival case history research and theory-based sampling. The two measuring instruments that were used are the items comprising the romance of leadership scale (RLS-C) and archival data. An exemplar of a Level 5 leader that has received ample press coverage, Bill Gates, was chosen. 201 articles from the following five sources and four years were collected and their content analysed: the Wall Street Journal, the Economist, Forbes, Fortune and Business Week in 1990, 1995, 1998 and 2006. The positive period included 1990 and 2006, and negative period 1995 and 1998. Once the 201 newspaper articles were randomly selected, the researcher analysed the content via 17 sub-categories. The eight sub-categories from the principles of Level 5 leadership were used as a criterion to find supportive evidence in the data for the Level 5 leadership theory. The researcher concurrently used the nine items from the RLS-C, which comprised nine sub-categories as a criterion to find supportive evidence in the data for romance of leadership theory. This research study resulted in a follower-centric, non-traditional recognition of leadership as it provided clear support for the romance of leadership theory. In the positive period, subcategory 1 and 5 dominated. Four trends were extracted: Ability, Image, Philanthropy and Innovation, constituting altogether 82 percent of the total propositions. Ability was the most salient trend. In the negative period, sub-categories 1 and 5 dominated. Five trends were extracted: Power, Image, Ability, Innovation and Competition constituting altogether, 85 percent of the total propositions. Power and Image were the most salient trends. Conceptual and methodological reasons for the findings were offered, followed by a discussion of the limitations and some theoretical and practical implications. Finally, future considerations for research were suggested.

Thesis advisor: Laura M. O'Dwyer / The purpose of this study was to inform future policy regarding school leadership practices through examining the relationship between reported decision making at the school level and student achievement. The study utilized a mixed methods design, and examined three main components. The first component involved a qualitative analysis of 14 countries' school leadership polices, as described in country background reports submitted to the Organization for Economic Cooperation and Development (OECD). The second component used hierarchical linear modeling (HLM) to examine the relationship between principal reported school decision making and student achievement in mathematics and reading on the Program for International Student Assessment (PISA) in the same 14 countries. The final component of this study connected the results from the policy analysis to the results of the HLM analysis to determine if there were patterns between a country's policies and the relationship between school-level decision making and student achievement. The study found that out of 14 countries included in the analysis, six countries were identified as having polices that were highly decentralized, three countries had policies that were highly centralized and five countries had policies that were somewhere in between the two extremes. The quantitative results showed that school-level decision making variables were limited in their utility as predictors of student achievement. Finally, an examination of the combined qualitative and quantitative results did not reveal any obvious patterns. However, the findings did highlight the importance of context in examining countries' policies and the relationship between leadership practices in the form of school decision making and student achievement. / Thesis (PhD) — Boston College, 2011. / Submitted to: Boston College. Lynch School of Education. / Discipline: Educational Research, Measurement, and Evaluation.