A methodological proposal for quality control of the soil moisture variable, measured in Colombian automatic agrometeorological stations.

Methodological criteria for data quality control with geophysical range and spectrum consistency were evaluated, establishing flags and quality indicators for soil moisture data records, in a range of depths between 10, 30, and 50 cm, from automatic agro-meteorological stations located in the most important agricultural regions of Colombia. Data for analysis were collected from 105 stations of the IDEAM network, in an observation window from 2001-2020. The results showed that 40.3% of the soil moisture data were of good quality, 12.9% were questionable due to spectrum flags, 14.3% were questionable due to geophysical range and 32% were erroneous because the values were not possible and/or missing. The depth closest to the surface had the highest number of quality flags, suggesting that the soil layer has the highest error detection rate associated with soil moisture condition recording; the most common quality flag was C02: "Soil moisture >60% & ≤100%", detected in 93% of the sensors, and the second most frequent flag was C01: "Soil moisture ≥0% & <3%". It was concluded that the proposed methodology provides highly satisfactory results in the detection of anomalous soil moisture records, in order to make adjustments to the environmental conditions of Colombia. [ABSTRACT FROM AUTHOR]