Atmospheric particles are important in climate, atmospheric chemistry, and human health studies. This is due to the face that they pose a risk to human health, with various adverse health effects for instance, respiratory inflammation and cardiovascular diseases. As such a deep understanding of spatial-temporal variation of particle number concentration under different meteorological conditions becomes crucial. Although studies on spatial and temporal variation of particles have been conducted in different regions of countries, multi-site studies including all kinds of areas (ranging from the most polluted to the clean air areas - e.g. mountainous site) using the long-term measurements in a country are rare. In this dissertation, the spatial and temporal variations of particle number size distributions (PNSD) due to meteorological influence in Germany were investigated. To obtain a comprehensive understanding of meteorological influences on spatio-temporal variations, the long-term measurements of particle number size distributions (approx. 500×103 data) based on hourly observations from German Ultrafine Aerosol Network (GUAN) over a span of four years (2011-2014) are analyzed. The particle number size distributions (PNSDs) of 14 sites in the size range between 20 and 800 nm were studied, having been classified into four individual site-types, viz. Roadside (RO), Urban background (UB), Rural background (RU) and Mountain sites (MS). The results illustrated that spatial variability in terms of ultrafine particles (UFPs with diameters < 100 nm) occurs at different time scales. The GUAN sites revealed distinct patterns of seasonal and annual periodicity with respect to sites with high influence from local emission sources while periodicity weakened in the less polluted sites. Findings were based on the meteorological parameters inherent to the respective sites. Moreover, the signatures of particle number size distributions important for interpreting PNSDs were studied. These signature size distributions (SSDs) were analyzed in relation to the meteorological parameters at different time scales. Two approaches of hierarchical- and K-means cluster analysis were used, and almost similar results were evident. The results of signature size distributions (SSDs) provided information on the behavior and patterns of cluster size distributions, as well as the influence of meteorological parameters on spatio-temporal variations. From the findings, clusters were sensitive enough to convey a similar signature in the specific site-types. This work endeavors to further knowledge, through the addition of a new site type, i.e. mountain site, to the catalogue of specific cluster signature size distribution. As a result, a more objective approach of classifying aerosols' characteristics is obtained. In this study ultrafine particles in the size range from 20 to 100 nm were observed at high III concentrations near the anthropogenic sources, and their small size could possibly explain their significant deposition in the deep lung and possible penetration through body tissue to the cardiovascular system. An investigation using machine learning models i.e. boosted regression trees (BRTs) and random forest (RF) algorithm, was carried out on the particle number concentrations for specific regions. These model proved useful in identifying and explaining relationships between explanatory variables, i.e. meteorological factors and ultrafine particles in developing training sites. Thereafter, each training site was compared with the other members of the same site-type (instead of making comparison with itself). By doing so, a more accurate conclusion about the homogeneity or heterogeneity of the sites could be obtained. Thus, each site-type had a suitable model, distinguishing this work from previous approaches. The relevance of meteorological factors on spatio-temporal variation and ultrafine particle number size distribution from the large data set covering four years were identified. Findings revealed that particle number concentrations (both on training and test sites) vary considerably in terms of space with respect to the specific regions, meteorology, and time. This study is presented in a monography form and was conducted in the framework of Ph.D thesis.