It is a well-known fact that an estimation of tropospheric gradients during GNSS data processing improves its quality mainly in terms of receiver position and Zenith Total Delay (ZTD) parameter. Tropospheric gradients are also essential for slant delays computation which are tested to be assimilated into numerical weather prediction models (NWM) and are used as input data for GNSS tomography. In our work we exploit the benchmark dataset collected within the COST GNSS4SWEC action which includes observations from 430 GNSS reference stations in central Europe for May and June 2013 and we focus on two main topics. Firstly, we provide an analysis of processing settings impact on estimated tropospheric gradients together with an evaluation of systematic effects in tropospheric gradients coming from gradient mapping function and observation elevation weighting. We used the Precise Point Positioning (PPP) processing in G-Nut/Tefnut software and delivered eight variants of solutions differing in gradient mapping function, elevation cut-off angle, GNSS constellation and real- time versus post-processing mode in order to assess their impact on tropospheric gradients. Moreover, tropospheric gradients from two NWMs were also derived to be used as reference products. Using statistical evaluation and visual inspection of tropospheric gradients maps we found a positive impact of lowering the cut-off elevation angle and using a combined GPS+GLONASS constellation. Moreover, all the post-processing variants of GNSS solution provided robust tropospheric gradients with a clear relation to real weather conditions. In regards of systematic effects in tropospheric gradients, e.g. the use of different gradient mapping function can lead to a difference in the north gradient component of up to 0.3 mm on a global scale due to latitudinal troposphere tilting. Secondly, we selected a set of weather events including various frontal system passages and local storm occurrences and visually studied tropospheric gradient maps in order to evaluate the potential of GNSS tropospheric gradients to capture them. Here we found a good correlation between GNSS tropospheric gradient fields and meteorological radar images showing instantaneous precipitation. Besides an ability to reliably capture large weather fronts the GNSS gradients from a dense network of stations were discovered to able to detect and track even a local storm life cycle. Our findings can potentially bring a new interest in GNSS tropospheric gradients usage in meteorological applications including nowcasting.