Validation of three - horizon soil maps based on different soil texture datasets for Denmark

In arable production and environmental planning, accurate soil maps are crucial for informed decision-making. However, creating precise digital soil maps is challenging due to regional and field spatial variations in soil texture caused by diverse geological processes. This study focuses on evaluating a Danish three-layer root zone (0–1.5 m) soil map (AC map). Over the last decade, the AC map has served as a foundational data layer in diverse applications, including enhancing national water resource models and simulation of nutrient dynamic for Denmark. The map includes soil texture, bulk density, content of organic matter, soil classes based on the Danish classification system (JB numbers), maximum rooting depth, and hydraulic parameters across three horizons. Validation is conducted by using two datasets: a dependent dataset (included in the production of the AC map), comprising data from the Danish textural database and Danish profile database, and an independent dataset from a national field trial database. The dependent dataset contained 37.948 points for the A horizon (0–30 cm), 5.892 for the B horizon (35–75 cm), and 1.136 for the C horizon (70–120 cm), while the independent dataset included 4.879 observation points for the A horizon. Validation employed the difference test (D value) and cumulative distribution functions (CDFs). Residual maps of clay content in the A horizon were prepared using kriging to analyze variability and D-value relationships. Results indicated low D-values, strong congruence between observed and predicted CDFs, and high predictive accuracy for both datasets in the A horizon. For the B horizon, the AC map exhibited higher D-values, such as 0.288 in geo-region 1 and 0.512 in geo-region 4, reflecting lower agreement with observed data, likely due to lower sample density and increased spatial variability. The C horizon showed the inconsistencies, with the highest D-values and weakest alignment to observed data, attributable to sparse sampling and the inherent challenges of predicting clay content. These findings confirm the AC map's reliability for environmental planning and soil management, particularly in the topsoil (0–30 cm).