Description: Carbon dioxide (CO2) and methane (CH4) are major greenhouse gases (GHG) and have been under constant monitoring for decades. Both gases have significantly increased in recent years due to anthropogenic activities. This has huge detrimental repercussions within natural systems including the warming of the planet. Although these GHG are extremely significant, there are also vast areas of study with little to no data in regards to emissions and budgets. These gaps are mainly within the aquatic regions (or the Land Ocean Aquatic Continuum (LOAC)). As a consequence, there can be large discrepancies between budget numbers and in turn, scaling and future predictions. In order to combine oceanographic and limnological methods this thesis presents a novel sensor package and show its application in multiple campaigns across the entire LOAC. The sensor set-up contained the oceanographic sensors HydroC CO2 FT (pCO2), HydroC CH4 FT (CH4), HydroFlash O2 (O2) and a thermosalinograph for temperature and conductivity measuring continuously, all simultaneously. We extensively mapped ocean to inland regions. The results first describe the processes to enable the set-up to be used across the LOAC boundary over 3 seasons. Extensive corrections were needed for the data to be fully appreciated for all salinities specifically in fresh inland waters. The data was then split between CO2 and CH4, where, in inland waters, further analyses were performed. The area of interest was the Danube Delta, which was found to be continuously supersaturated in regards to CH4 and fluctuating between a source and sink for CO2. Extraction of TA was possible, using the sensors continuous data by applying a simple model. In this extraction and the continuous data, large spatial-variability was observed and further analysed allowed for diel cycle extractions, which are usually disregarded in budgets and measurements. In channels, CH4 concentrations and fluxes were found to potentially be underestimated by up to +25% and +20% respectively when not including a full diel cycle. In lakes however, we found the opposite, with an overestimation in concentration and fluxes (+3.3% and +4.2%) when not considering the diel cycle, although this greatly depends on time of the sampling.