Homogenised input series for generating gridded data, does it matter?

When assessing long term climate trends and variability it is important to analyse data that are not disturbed by external factors that might lead to misleading trends. It is also urgent to analyse consistent series that cover the entire time period. Since observation networks are continuously changing not too many complete series are available for a centennial long analysis.  Gridded climate monitoring data are often based on all available data for each timestep, which might disturb the spatial and temporal consistency. In order to investigate such impacts we have constructed a long term gridded dataset based on homogenised input data.  165 raw temperature series and 323 raw precipitation series fr Norway covering the entire period 1901-2020 were constructed by an extraction from a gridded data set of monthly climate anomalies, which was based on all available observations. The constructed series were homogenised applying the automatic procedure in Climatol.  New gridded data sets were established applying the homogenised data series as input. These data sets were compared to a similar gridded dataset based on the raw non-homogenized data series. Analysis of the regional time series of the datasets shows similar temporal variability and trends. The gridded data based on the homogenised data shows less local spatial variability than those based on the raw data series. This means that anomalies due to inhomogeneous data are reduced, and that false local trends are removed. The construction and homogenization of climate data series therefore provides more consistent data for gridding, and leads to better and more reliable gridded data sets for assessing spatio-temporal climate trends.