In preparation for, and for decision support during, CBRN (chemical, biological, radiological and nuclear) emergencies it is essential to know how such an event would turn out, so that one can prepare a possible evacuation. Afterwards it might be good to know how to backtrack and see what caused the emergency, and in the case of e.g. a gas leak, where did it begin? The Swedish Defence Research Agency (FOI) develops models for such scenarios. In this thesis FOI's model, "The Nordic Urban Dispersion model" (NUD), has been studied. The system of equations set up by this model was originally solved using Intel's PARDISO solver, which is a direct solver. An evaluation on how an iterative multigrid method would work to solve the system has been done in this thesis. The wind model is a mass-consistent model which sets up a diagnostic initial wind field. The final wind field is later minimized under the constraint of the continuity equation. The minimization problem is solved using Lagrange multipliers and the system turns into a Poisson-like problem. The iterative algebraic multigrid solver (AMG) which has been evaluated had difficulties solving the problem of an asymmetric system matrix generated by NUD. The AMG solver was then tried on a symmetric discrete Poisson problem instead, and the solution turns out to be the same as for the PARDISO solver. A comparison was made between the AMG and PARDISO solver, and for the discrete Poisson case the AMG solver turned out on top for both larger system size and less computational time. To try out the solvers for the original NUD case a modification of the boundary conditions was made to make the system matrix symmetric. This modification turns the problem into a mathematical problem rather than a physical one, as the wind fields generated are not physically correct. For this modified case both the solvers get the same solution in essentially the same computational time. A method of how to in the future solve the original (asymmetric) problem, by modifying the discretization of the boundary conditions, has been discussed.

Simulations of the wind climate in complex terrain may be useful in many cases, e.g. for wind energy mapping. In this study a new mass-consistent model (MCM), the λ-model, was developed and the ability of the model was examined. In the model an initial wind field is adjusted to fulfill the requirement of being non-divergent at all points. The advance of the λ- model compared with previous MCM:s is the use of a terrain-influenced coordinate system. Except the wind field, the model parameters include constants α, one for each direction. Those constants have no obvious physical meaning and have to be determined empirically. To determine the ability and quality of the λ-model, the results were compared with results from the mesoscale MIUU-model. Firstly, comparisons were made for a Gauss-shaped hill, to find situations which are not caught by the λ-model, e.g. wakes and thermal effects. During daytime the results from the λ-model were good but the model fails during nighttime. From the comparisons between the models the importance of the α-constants were studied. Secondly, comparisons between the models were made for real terrain. Wind data from the MIUU-model with resolution 5 km was used as input data and was interpolated to a 1 km grid and made non-divergent by the λ-model. To study the quality of the results, they were compared with simulations from the MIUU-model with resolution 1 km. The results are quite accurate, after adjusting for a difference in mean wind speed between MIUU-model runs on 1km and 5 km resolution. Good results from the λ-model were reached if a climate average wind speed was calculated from several simulations with different wind directions. Especially if the mean wind speed for the domain in the λ-model was modified to the same level as in the MIUU 1 km. The λ-model may be a useful tool as the results were found to be reasonable good for many cases. But the user must be aware of situations when the model fails. Future studies could be done to investigate if the λ-model is useable for resolutions down to 100 meters. / Modellering av vindklimat i komplex terräng är användbart i många sammanhang, t ex vid vindkartering för vindenergi. I den här studien utvecklas och undersöks användbarheten av en sk. Mass-Consistent Model, λ-modellen. Modellen bygger på att ett initialt vindfält justeras för att uppfylla kontinuitetsekvationen i alla punkter. För att göra vindfältet divergensfritt används en metod som bygger på variationskalkyl. Fördelen med denna nya modell jämfört med tidigare är användandet av ett terränginfluerat koordinatsystem. I teorin för λ-modellen införs en parameter α. Då denna inte har någon självklar fysikalisk betydelse behöver den bestämmas empiriskt.   För att undersöka kvalitén hos λ-modellen gjordes jämförelser med den mesoskaliga MIUU-modellen. Det första steget var att jämföra körningar över en Gaussformad kulle, detta för att jämföra modellerna och finna situationer som λ-modellen inte löser upp. Exempel på sådana är termiska effekter och vakar. Resultaten under dagtid var bra medan under nattetid var det stora skillnader mellan modellerna. Utifrån resultaten kunde betydelsen av α-parametern studeras.   Nästa steg var att jämföra med verklig terräng. Detta gjordes för ett område i Norrbotten. Här användes vinddata från MIUU-modellen med upplösning 5 km som indata för att beräkna vinden på en skala 1 km. För att undersöka kvalitén hos λ-modellen användes data från MIUU-modellen med upplösning 1 km som jämförelse. Resultaten avseende vindvariationerna i terrängen är tillfredställande, dock med något för höga vindhastigheter i λ-modellen. Detta visade sig bero på för högre medelvind i MIUU 5 km än i MIUU 1 km. Jämförelse mellan modellerna gjordes även för Suorva-dalen i Lappland vilken omges av bergig terräng. Resultaten här var sämre avseende medelvindarna, men med bättre resultat avseende vindriktningarna.   Bra resultat för λ-modellen nåddes då resultat från flera simuleringar slogs samman till ett medelvärde. Framförallt blev resultatet bra då medelvinden justerades till samma nivå som MIUU 1 km.   Sammanfattningsvis kan sägas att resultaten från λ-modellen är rimliga i många situationer men att det är viktigt att veta i vilka situationer den inte fungerar. Framtida undersökningar bör göras för att undersöka om modellen är användbar för upplösningar ner till ca 100 meter.

La méthode de projection et l'approche variationnelle de Sasaki sont deux techniques permettant d'obtenir un champ vectoriel à divergence nulle à partir d'un champ initial quelconque. Pour une vitesse d'un vent en haute altitude, un champ de vitesse sur une grille décalée est généré au-dessus d'une topographie donnée par une fonction analytique. L'approche cartésienne nommée Embedded Boundary Method est utilisée pour résoudre une équation de Poisson découlant de la projection sur un domaine irrégulier avec des conditions aux limites mixtes. La solution obtenue permet de corriger le champ initial afin d'obtenir un champ respectant la loi de conservation de la masse et prenant également en compte les effets dûs à la géométrie du terrain. Le champ de vitesse ainsi généré permettra de propager un feu de forêt sur la topographie à l'aide de la méthode iso-niveaux. L'algorithme est décrit pour le cas en deux et trois dimensions et des tests de convergence sont effectués. / The Projection method and Sasaki's variational technique are two methods allowing one to extract a divergence-free vector field from any vector field. From a high altitude wind speed, a velocity field is generated on a staggered grid over a topography given by an analytical function. The Cartesian grid Embedded Boundary method is used for solving a Poisson equation, obtained from the projection, on an irregular domain with mixed boundary conditions. The solution of this equation gives the correction for the initial velocity field to make it such that it satisfies the conservation of mass and takes into account the effects of the terrain. The incompressible velocity field will be used to spread a wildfire over the topography with the Level Set Method. The algorithm is described for the two and three dimensional cases and convergence tests are done.

La méthode de projection et l'approche variationnelle de Sasaki sont deux techniques permettant d'obtenir un champ vectoriel à divergence nulle à partir d'un champ initial quelconque. Pour une vitesse d'un vent en haute altitude, un champ de vitesse sur une grille décalée est généré au-dessus d'une topographie donnée par une fonction analytique. L'approche cartésienne nommée Embedded Boundary Method est utilisée pour résoudre une équation de Poisson découlant de la projection sur un domaine irrégulier avec des conditions aux limites mixtes. La solution obtenue permet de corriger le champ initial afin d'obtenir un champ respectant la loi de conservation de la masse et prenant également en compte les effets dûs à la géométrie du terrain. Le champ de vitesse ainsi généré permettra de propager un feu de forêt sur la topographie à l'aide de la méthode iso-niveaux. L'algorithme est décrit pour le cas en deux et trois dimensions et des tests de convergence sont effectués. / The Projection method and Sasaki's variational technique are two methods allowing one to extract a divergence-free vector field from any vector field. From a high altitude wind speed, a velocity field is generated on a staggered grid over a topography given by an analytical function. The Cartesian grid Embedded Boundary method is used for solving a Poisson equation, obtained from the projection, on an irregular domain with mixed boundary conditions. The solution of this equation gives the correction for the initial velocity field to make it such that it satisfies the conservation of mass and takes into account the effects of the terrain. The incompressible velocity field will be used to spread a wildfire over the topography with the Level Set Method. The algorithm is described for the two and three dimensional cases and convergence tests are done.