Abstract
The standard normal homogeneity test has been applied to 165 Norwegian precipitation series of 75 years or more. Of these series, 50 were found to be homogeneous, while 79 became homogeneous after being adjusted for a single inhomogeneity. Almost 50% of the inhomogeneities were caused by relocation of the precipitation gauge. Other reasons for inhomogeneities were changes in the immediate environment (trees, buildings) and changes in instruments (windshield, new gauge type). About 80% of the inhomogeneities could be traced to information in the station history flies. Most of the inhomogeneities caused by changes in instrumentation led to increased gauge catch. There was a similar tendency for inhomogeneities caused by changes in environment. Consequently, trend studies based on groups of untested series may give dubious trends. Homogeneity testing and station history archives are essential tools for finding the real trends and fluctuations in precipitation series.
