A numerical modell has been developed at the Department of Meteorology, Uppsala University. The purpose has been to calculate streamline displacements for linear, hydrostatic flow over a three-dimensional hill and to visualize the characteristics of the flow. The model simulates the vertical and lateral displacements and also velocity perturbation fields for flow approaching the hill at constant mean wind and constant stability. The simulations show the lateral deflections, as the air tries to avoid the hill, the horizontal deceleration and ascent upstream and the horizontal acceleration on the lee-slope. Divergence at the mountain top is compensated by descent of air from aloft. Gravity waves propagate upwards and partly outwards along parabolas. It is also shown how linear theory fails to be valid when the Froude number Fr → 1+.

Meteorologiska observationer har gjorts vid Observatoriekullen i Stockholm sedan 1754 och pågår fortfarande. Termometerns placering har ändrats vid två tillfällen, år 1875 och år 1960. Före år 1875 är dokumentationen bristfällig. Den enda informationen som finns att tillgå är att termometern var placerad på andra våningen på observatoriets norra sidan och att den hängde ute i friska luften samt att den var väl skyddad från solen. I denna studie har gjorts ett försök att återskapa observationsplatsen som den var före år 1875 och jämföra den med den plats där SMHI idag utför sina mätningar. En försöksuppställning med två resistansgivare monterades på en plats som troligen motsvarar den plats där observationerna ursprungligen utfördes. En av termometrarna försågs med ett enkelt skuggskydd och den andra försågs med ett modernt strålningsskydd. Dessa två givare jämfördes med en givare placerad i SMHIs bur. Mätningarna pågick under perioden juni 1995—maj 1997. Temperaturdifferenserna kategoriserades efter molnmängd och vindstyrka. De största differenserna erhölls under dagar med klart väder och låga vindhastigheter. De minsta skillnaderna registrerades under dagar med mulet väder och höga vindhastigheter. Månadsmedelvärden beräknade med Hambergs och Ekholm—Modéns formler jämfördes också. Den förra användes under perioden fram till 1858. Den senare är den som används idag. I medeltal var skillnaden mellan givarna vid den gamla mätplatsen, beräknade med Hambergs formel, och givaren vid SMHIs nuvarande plats, beräknad med Ekholm—Modéns formel cirka 0,1º. För enskilda månader kunde dock differensen uppgå till 0,5º, vilken ungefär är densamma som man räknar med att temperaturen har stigit, under de sista 130 åren, som en följd av urbaniseringseffekten.

Aircraft measurements are analysed from the “First Lagrangian” of the Atlantic Stratocumulus Transition Experiment (ASTEX) from south east of the Azores Islands. In this experiment, Lagrangian strategy was used and the marine air mass, that advected southward, was followed during 12 to 14 June 1992. During the experiment, the stratocumulus clouds transitioned into thin and broken stratocumulus with cumulus cloud penetrating from below. To characterise the vertical structure in the marine boundary layer the buoyancy fluxes, the variances, the turbulent kinetic energy, the momentum fluxes and humidity fluxes were examined. The buoyancy flux profiles were used to discover the decoupling of the stratocumulus and the sub-cloud layer. Turbulence analysis for all five flights shows that the cloud layer were decoupled from the underlying layer. In the cloud layer the buoyancy production due to longwave radiative cooling at cloud top, was the main source for driving the turbulence. In the sub-cloud layer, the variances indicate wind shear to be the main generator of turbulence for the first two days. Then, as sea surface temperature increases, buoyancy produced turbulence was more pronounced. The u-, v- and w-spectra and cospectra of wθ and uw give insight into the typical eddy sizes, and how the peak wavelengths vary with height. The peak wavelengths in sub-cloud and cloud layer were larger than layer depth and u- and v-spectral peak wavelengths often larger than the peak wavelength from w-spectra. While peak wavelengths in the sub-cloud layer vary with the height above the surface, they are approximately invariant with height in the cloud layer.

The effect that the Polar stratospheric clouds (PSCs) have on the destruction of ozone in the polar vortex is well known. The study of these clouds is therefore highly relevant and important. A tool that can be used to gather parameters needed to describe the PSC is a lidar (Light Detection and Ranging). The Swedish Institute of Space Physics (IRF) uses such a lidar to observe PSCs above Kiruna in northern Sweden. The lidar works by emitting polarised light and then collecting the scattered light. Knowing the time between when the light was emitted and the backscattered light was collected one can then determine at which altitude the scattering happened. The time and the backscattered light can then be used to derive the useful parameters needed to create a climatology of the PSCs. When creating a climatology, the different types of PSCs (type Ia, Ib and II) have to be classified. Each of the different types is classified by looking at the PSCs parallel and perpendicular backscatter ratio as well as the depolarisation ratio. The classification can then be used to understand the formation and composition of the cloud. To process the data gathered from the lidar a script in MATLAB is developed. The script is automatically going through the data and pinpointing the boundaries and interpolate the data, then using this to calculate the backscatter and depolarisation ratio. The script runs for all the available IRF observations between 2008 and 2018. The ratios and boundaries calculated from all the observations are then used to create a climatology. Comparing the climatology to the climatology by Blum et. al \cite{Psct} shows similar results. Future work should be done on the MATLAB program to inprove finding parameters from observation where the data includes more noise than the program can currently handle. This enables the retrieval of more trusty parameters and a more precise climatology.

The Arctic Summer Cloud Ocean Study expedition took place during the autumn of 2008 and brought back water and air samples. One theory was that marine particles were shot into the atmosphere by bubble bursting and, while in the atmosphere, acted as cloud condensation nuclei. Particles collected from the subsurface water, surface microlayer and just above the surface had their topography and morphology analyzed using a scanning electron microscope. Due to a lack of EDS analysis it's impossible to say for sure if the particles found were the same found in previous studies, just that it is highly likely that they are. No evidence against the marine particles being able to act as cloud condensation nuclei was found but the data gathered was not sufficient to strengthen the theory either. The scanning electron microscope was also used in a pedagogical study, analyzing how operators with different knowledge and prior experience interact with the microscope's images. These results as well as knowledge gained from literature studies were used to improve a high school experiment regarding centripetal acceleration. The main issue with the experiment was that what the students learned performing the experiment was not the same as the theoretical models the teachers wished the students to learn. The reason for this was because the experimental equipment lacked the centripetal model's external representations such as force arrows. If a simulator would be incorporated into the lab centripetal acceleration representations could be visualized and a clearer connection between experiment and theory could be made.

Wind atlases are interactive databases used extensively for preliminary wind resource assessment aimed at optimal wind power development. The selection and processing of wind atlas data are important and will determine the quality of the subsequent WRA (Wind Resource Assessment) phases. This study aims to compare, process, and interpret different wind atlases in determining effective wind resource analysis for 14 different sites in Sweden. The wind atlas data is accessed, parsed, and compared using multiple methodologies to ensure a fair comparison. Several variables like wind speed, wind direction, temperature, production values, etc. are analysed and compared amongst the wind atlas data. The corrected versions of the data are drawn out to conclude and determine how well those databases represent the realistic conditions in a forested site. The data is also compared and validated with measurement data. The results indicate several layers of insights into the wind resource regime. The preliminary results from 14 sites show that New European Wind Atlas (NEWA) and Global Wind Atlas (GWA) agreed sufficiently on the subject of the wind conditions for most of the sites. Methodologies like downscaling and long-term correction introduced to the wind atlas data provided more insights and proved that the database could expand and include more complex wind conditions in the coming years. Furthermore, the comparison and analysis of wind atlas data along with measurement data revealed that NEWA aligned the most to the actual wind conditions for a site. Several uncertainties are considered but more validation and thorough insights in the future could determine and give more affirmations to the idea put forward. Further studies could be conducted along with the inclusion of additional layers, variables, and databases to the wind atlases.

Mätningar av gränsskikt som utförts i Marsta (MAO) under tidsperioden 1 juli 1994 till 28 februari 1995 har analyserats. Mätningarna, som genomfördes på en 30 meter hög mast, omfattar bland annat vindhastighet, vindriktning och temperatur på 6 nivåer. Långvågig in- och utstrålning uppmättes på en höjd, vilket även gällde turbulenta fluktuationer av vindens 3 komponenter och temperatur. Den här redovisade undersökningen inriktades på studier av stabila gränsskikt. Dels studerades turbulensen samt hur temperaturavtagandet vid bildningen av ett stabilt gränsskikt varierar med tiden. Även specialfallet, när ett neutralt skikt bildas vid markytan i det stabila gränsskiktet mitt i natten, studerades. För undersökning av turbulensen har jämförelser mellan dessa mätdata och andra undersökningars mätdata genomförts. Jämförelsen har gjorts med hjälp av Monin—Obukhovs similaritetsteori vilket innebär att man studerar hur de dimensionslösa standardavvikelsern för u, v, w och T och den dimensionslösa temperaturgradienten beror av z/L. Resultaten visade att överensstämmelsen mellan denna och andra undersökningar var relativt god. Genom numeriska beräkningar har man tidigare funnit att temperaturen nära marken minskar med roten ur tiden de första 3-4 timmarna, då skiktningen blir stabil. Detta gäller vid konstant långvågsutstrålning. I den här undersökningen uppgick motsvarande tidsperiod till 45 minuter. En ekvation som beskriver hur det neutrala skiktets höjd växer har härletts. Denna ekvations giltighet har även testats. Resultatet var inte entydigt.

Measurements collected at Cabauw, the Netherlands, during the period November 1996 to May 1997, have been evaluated. The study was focused on the calculation of turbulence characteristics and profiles of wind speed and temperature under stable and near neutral conditions in the atmospheric surface-layer. Results from a 1-dimensional second-order closure model was compared with the outcome from the present observational data. The Monin-Obukhov similarity was applied for calculating the non-dimensional wind (Φm) and temperature gradients (Φh), the obtained result are: Φm = 0.9 + 4.4z/L and Φh = 0.9 + 3.6z/L. We suggest to replace the length scale kz, in the length scale of the definition of the Φ-functions, with the integral length scale, Λw, since Λw is more related to the length scale of the most energetic turbulent motions. On basis of the measurement campaigns performed, it was not possible to verify unique functions for these non-dimensional profiles, and therefore it is suggested that another length scale besides kz must be of importance. Furthermore, the balance between shear production, viscous dissipation and the buoyancy in the turbulent kinetic energy budget were investigated. The values of the non-dimensional standard deviations of horizontal and vertical velocities are constant with regard to stability and in good agreement with figures reported in the literature. By using a 1-dimensional second-order closure model it has not been possible to describe the turbulence characteristics of the stable atmospheric surface-layer. Three different combinations of constants are shown in this report, and they indicate problems that occur when comparing closure model results with experimental data from a stable stratified atmosphere.

This thesis will examine the local effects of land cover on icing forecasts. In Paper I, a single column model was used to test the sensitivity of icing forecasts to land cover fraction. Here, the ice accretion forecast was found to be highly sensitive to the wind magnitude response to the surface roughness. Diabatic effects related to the surface albedo played a secondary role, significant in cases with strong solar irradiance. Paper II examined the impact of 2-dimensional patterns of land cover heterogeneity on the effective surface roughness and blending height using large eddy simulation over a diurnal cycle of solar irradiance. The blending height--or the elevation at which the atmospheric response to the underlying land cover becomes horizontally homogeneous--has been proposed as a guide for coupling numerical weather models to surface parameterizations. In stable conditions, when the atmospheric boundary layer height was shallow, the blending height over surfaces with large heterogeneity length scale was found to be much lower than that of analytical models from previous studies. A new formula for a dynamic blending height was proposed taking this effect into account. The effective surface roughness was found to decrease with increasing land cover heterogeneity. The wind power response from an idealized wind turbine with 80-meter hub height to the effective surface roughness was shown, with a positive response in wind power with increasing land cover heterogeneity. The wind power response was smaller and less systematic with wind turbines extending above the blending height, further highlighting the utility of an accurate formulation for this variable. / I den här avhandligen undersöks de lokala effekterna av landtäcke på isbildningsprognoser. I artikel I användes en enkolumns-modell för att testa känsligheten hos isbildningsprognoser för olika andelar landtäcke. Här visade sig isbildningsprognosen vara mycket känslig för ytans påverkan på vindstyrkan. Diabatiska effekter relaterade till ytalbedo spelade en sekundär roll, speciellt i fall med stark solinstrålning. Artikel II undersökte hur  tvådimensionella mönster i landtäckets heterogenitet påverkade den effektiva ytjämnheten och blandningshöjden med hjälp av simuleringar av en stor virvel under en dygnscykel av solinstrålning. Blandningshöjden - eller den höjd vid vilken atmosfärens respons på det underliggande marktäcket blir horisontellt homogent - har föreslagits fungera som  guide för  hur man kan koppla numeriska vädermodeller till ytparameteriseringar. Under stabila förhållanden, när den atmosfäriska gränsskiktshöjden var låg, visade sig blandningshöjden över ytor med stor heterogenitetslängdskala vara mycket lägre än den från analytiska modeller från tidigare studier. En ny formel för att beräkna dynamisk blandningshöjd föreslogs med hänsyn till denna effekt. Den effektiva ytjämnheten visade sig minska med ökande heterogenitet i landtäcket. Responsen med avseende på den effektiva ytjämnheten visades för en ideal vindturbin med 80 meters navhöjd , med en ökning i vindkraftsproduktion med ökande heterogenitet i landtäcket. Påverkan på produktionen var både mindre i allmänhet och även mindre systematisk för vindkraftverk som sträckte sig över blandningshöjden, vilket ytterligare understryker behovet av en exakt formulering av denna variabel.

Dimma är ett komplext småskaligt väderfenomen som gäckat meteorologerna ända sedan väderprognosernas start och varit en bidragande orsak till att många mist livet inte minst inom flyget. Än idag är prognostisering av dimma mycket svårt i synnerhet strålningsdimma. Studier (COST, 2007) har visat att för att en modell ska kunna fånga formation av dimma krävs det i vissa fall att den skall kunna lösa upp inversionsskikt på 10m.Därtill uppgår osäkerheten i många fall för numeriska modeller vid fastställande av sikten i dimma till 50 %. Då kan en statistisk kartläggning specifikt för varje dimtyp utgöra ett bra komplement till modellerna för att få en uppfattning om exempelvis sikten i dimman. I den här studien användes observationer gjorda med en halvtimmes mellanrum från Arlanda under perioden 1993-07-01 – 2009-11-03.I denna kartläggning identifierades de fyra dimtyperna: strålningsdimma, nederbördsdimma, advektionsdimma samt dimma orsakad av en sänkning av molnbasen till marknivå (molndimma). Här framkom att strålningsdimman var den vanligaste dimtypen (43,6%) och mest frekvent under sommaren och hösten, till skillnad från nederbördsdimma och advektionsdimma som var vanligast under vinterhalvåret. För strålningsdimman fann man också de sämsta siktvärdena följt av molndimma. Det framkom också att frekvensen av de fall där strålningsdimman täcker hela flygplatsen avsevärt minskat sedan början på 2000-talet, detta samtidigt som utbyggnaden av flygplatsen stod klar.Vid studierna av isdimma och vattendimma visade det sig att den lägsta sikten för vattendimmor uppkom då temperaturen och den specifika fuktigheten var som högst. Medan för isdimma fann man de lägsta siktvärdena vid temperaturer runt -20 °C / Since the beginning of weather forecasting fog has caused many meteorologists severe problems, and in aviation it has been one of the main reasons for several crashes. Even today with the latest numerical models, predicting fog is a difficult task. New studies have shown that to predict fog the model must be able to resolve inversion layers down to 10 m. Even then the precision in determination of the visibility inside the fog is down to 50 % (Gultepe, 2007). In those cases a statistical mapping of fog at a certain place could be a good complement to the numerical models in determining time of formation and visibility. In this study I have used METAR observations made at Arlanda airport from 1993-07-01 to 2009-11-03, which are made every half an hour. In the mapping Radiation fog, Advection fog, Precipitation fog and fog caused by lowering of clouds (cbl. fog) was sorted out. I also looked into the behavior of ice fog vs. water fog.The results showed that radiation fog was the most common fog type (43,6 %) and had the lowest rate of visibility followed by cbl. fog. Radiation fog also had its highest frequency on summer and autumn, while advection fog and precipitation fog was concentrated into wintertime. Additionally the study showed that the frequency of the times when radiation fog covers the whole airport started to decrease at the beginning of the 21st century. Coinciding with when the expansion of the airport was finished including a new runway. In the comparison between water fog and ice fog it was clear that the distribution of visibility was similar. The lowest visibilities in water fog was found at the highest temperatures and highest specific liquid water content while in ice fog the lowest visibilities where found for temperatures in the neighborhood of - 20 °C. For temperatures below - 20 °C there were no observations of fog.