Abstract: Based on the existing daily homogenized datasets of Tianjin and the surface meteorological observation data A file, the metadata related to the relocation and instrument changes at 14 meteorological stations during the period of 2012 to 2024 were supplemented as detailed as possible, and using the method of combination of empirical decision, without reference series and Quantile-Matching (QM) adjustment, the homogenized daily datasets version 2.0 of essential climate variables in Tianjin covering 1951-2024 was developed. Results indicate that: (1) Except for the elements of sea level pressure and precipitation, the series of the other elements have statistically significant breakpoints (at 1% level). Among them, the maximum temperature and 2-minute wind speed are more obviously affected by non-homogenization factors than the others, which have the stations with breakpoints reaching 71.4% and 85.7%, respectively. The influence reasons for these two elements are mainly instrument changes and the combined effects of relocation and instrument changes; (2) The QM adjustments show that instrument changes have caused systematic overestimation of temperature observation data, and the average adjustment for the daily maximum temperature is approximately 1.2°C. The overall adjustment for 2-minute wind speed is mainly positive, with an average amount of 0.3 m/s, but the offshore station (54646) has a distinct negative adjustment due to the excessive response demonstrated by the old instrument; (3) The analyses of the statistical values before and after adjustment reveal that the non-climatic disturbances in the climate observation series are visibly weakened by QM adjustment. False fluctuations and trend jumps of the maximum temperature have been effectively suppressed, and the long-term downward trend of the 2-minute wind speed has been further highlighted. Additionally, the climate physical risk index (CPRI) based on the adjusted data is more in line with the observed facts caused by the frequent occurrence of extreme climate events in the current and recent decades. Therefore, to some extent, the importance of observation data homogenization in improving the reliability of regional climate monitoring and supporting scientific assessment for climate adaptation can be reflected in this study.