Globally, snow influences Earth and its ecosystems in several ways by having a significant impact on, e.g., climate and weather, Earth radiation balance, hydrology, and societal infrastructures. In mountainous regions and at high latitudes snowfall is vital in providing freshwater resources by accumulating water within the snowpack and releasing the water during the warm summer season. Snowfall also has an impact on transportation services, both in aviation and road maintenance. Remote sensing instrumentation, such as radars and radiometers, provide the needed temporal and spatial coverage for monitoring precipitation globally and on regional scales. In microwave remote sensing, the quantitative precipitation estimation is based on the assumed relations between the electromagnetic and physical properties of hydrometeors. To determine these relations for solid winter precipitation is challenging. Snow particles have an irregular structure, and their properties evolve continuously due to microphysical processes that take place aloft. Hence also the scattering properties, which are dependent on the size, shape, and dielectric permittivity of the hydrometeors, are changing. In this thesis, the microphysical properties of snowfall are studied with ground-based measurements, and the changes in prevailing snow particle characteristics are linked to remote sensing observations. Detailed ground observations from heavily rimed snow particles to openstructured low-density snowflakes are shown to be connected to collocated triple-frequency signatures. As a part of this work, two methods are implemented to retrieve mass estimates for an ensemble of snow particles combining observations of a video-disdrometer and a precipitation gauge. The changes in the retrieved mass-dimensional relations are shown to correspond to microphysical growth processes. The dependence of the C-band weather radar observations on the microphysical properties of snow is investigated and parametrized. The results apply to improve the accuracy of the radar-based snowfall estimation, and the developed methodology also provides uncertainties of the estimates. Furthermore, the created data set is utilized to validate space-borne snowfall measurements. This work demonstrates that the C-band weather radar signal propagating through a low melting layer can significantly be attenuated by the melting snow particles. The expected modeled attenuation is parametrized according to microphysical properties of snow at the top of the melting layer.

Cumulonimbus (Cb) clouds form a serious threat to aviation as they can produce severe weather hazards. Therefore, it is important to detect Cb clouds as well as possible. Finnish Meteorological Institute (FMI) provides aeronautical meteorological services in Finland, including METeorological Aerodrome Report (METAR). METAR describes weather at the aerodrome and its vicinity. Significant weather is reported in METARs, and therefore Cb clouds must be included in it. At Helsinki-Vantaa METARs are done manually by human observer. Sometimes Cb detection can be more difficult, for example, when it is dark, and it is also expensive to have human observers working around the clock all year round. Therefore, automation of Cb detection is a topical matter. FMI is applying an algorithm that uses weather radar observations to detect Cb clouds. This thesis studies how well the algorithm can detect Cb clouds compared to manual observations. The dataset used in this thesis contains summer months (June, July and August) from 2016 to 2020. Various verification scores can be calculated to analyse the results. In addition, daytime and night-time differences are calculated as well as different years and months are compared together. The results show that the algorithm is not adequate to replace human observers at Helsinki-Vantaa. However, the algorithm could be improved, for instance, by adding satellite observations to improve detection accuracy.

Observations have been and are an important part of today's meteorological developments. Surface observations are very useful as they are, providing weather information for a point location. ough they do not give much information, if any, on what happens between the stations across a larger area. With models one can create an analysis of the meteorological situation, i.e. calculate and estimate what happens between these fixed observation points. Remote-sensing data, such as radar and satellite, are being processed and the output is given over a domain as an analysed product of their measurements. For example, radar gives a plot of where the rain is located, i.e. an analysis of the current precipitation. With a series of radar images, a human (subjectively) or a computer objectively) can process this information to estimate where the rain will move and be located within the next few minutes (even hours), i.e. a short forecast also called "nowcast". is applies to some extent also for other observations, such as satellite data (cloud propagation). But for most quantities (such as temperature, wind, etc) it is significantly harder to make such a nowcast, since these are influenced by many other factors and there is no linear development of them. Therefore, there are forecast models that solve physical and dynamic equations, so that one can estimate the future weather for the coming hours and days. A prerequisite for generating a forecast of high quality is to capture the initial weather conditions as best as possible. This is done using observations and they are introduced into the forecast model through different techniques, where the model creates its own analysis as the initial step. There remain problems since forecast models often are affected by physical disagreements, as the dynamic conditions are not in balance. This results in the model having a spin-up effect, where the meteorological quantities are not yet in balance with each other and the resulting weather conditions are not always reliable during the first hours. Hence, a lot of research is spent on how to reduce this spin-up effect and on the use of nowcast models, in order to deliver the best model results for the first few hours of the forecast period. In this dissertation, the research work has been to improve the meteorological analysis, algorithms and functionality, using the Local Analysis and Prediction System (LAPS) model. Different kinds of observations were used and their interdependencies have been studied, in order to combine and merge information from variousinstruments. Primarily focus has been to improve the estimation of precipitation accumulation and meteorological quantities that affect wind energy. The LAPS developments have been used for several end-users and nowcasting applications, and experimentally as initial conditions for forecast modelling. The studies have been concentrated on Finland and nearby sea areas, with the available datasets for this domain. By combining surface-station measurements, radar and lightning information, one can improve the precipitation-amount estimations. The use of lightning data further improves the estimates and gives the advantage of having additional data outside radar coverage, which can potentially be very useful for example over sea areas. In addition, the improved LAPS analyses (cloud-related quantities) and a newly developed model (LOWICE), calculating the electricity production during wintertime (taking into account the icing of wind turbine rotor blades which reduces efficiency), have shown good results.