Modelling effective soil depth at field scale from soil sensors and geomorphometric indices

Abstract

The effective soil depth (ESD) affects both dynamic of hydrology and plant growth. In the southeast of Buenos Aires province, the presence of petrocalcic horizon constitutes a limitation to ESD. The aim of this study was to develop a statistic model to predict spatial patterns of ESD using apparent electrical conductivity at two depths: 0-30 (ECa_30) and 0-90 (ECa_90) and geomorphometric indices. To do this, a Random Forest (RF) analysis was applied. RF was able to select those variables according to their predictive potential for ESD. In that order, ECa_90, catchment slope, elevation and ECa_30 had main prediction importance. For validating purposes, 3035 ESD measurements were carried out, in five fields. ECa and ESD values showed complex spatial pattern at short distances. RF parameters with lowest error (OOBerror) were calibrated. RF model simplified which uses main predictors had a similar predictive development to it uses all predictors. Furthermore, RF model simplified had the ability to delineate similar pattern to those obtained from in situ measure of ESD in all fields. In general, RF was an effective method and easy to work. However, further studies are needed which add other types of variables importance calculation, greater number of fields and test other predictors in order to improve these results.

Mapping soil depth in southern pampas Argentina using ancillary data and statistical learning

“…The variability of soil depth accounted for by ECa and terrain attributes and OOB accuracy achieved by the model trained with the full dataset was 66% and 21.8 cm, respectively (Table 2). These results agree with those reported in the literature (Castro-Franco et al, 2017;Domenech et al, 2017;Tol et al, 2013). Compared with the overall mean of soil depth as a benchmark predictor, the model based on these features represents an accuracy gain of 14%; although certainly modest, the ability to predict this attribute based on readily available data would represent a good alternative over time-consuming soil depth surveys.…”

“…However, for Paleudolls, the variables related to EC90 followed EC30 in importance, and landscape attributes were not included within the top 10 most important variables. The importance of EC30, EC90, and elevation as predictors agrees with results reported for similar soils (Castro-Franco et al, 2017;Domenech et al, 2017;Paggi et al, 2013). Although the ECa is negatively related to soil depth in both soils, the reasons behind them may be slightly different.…”

“…However, the ability to generalize these patterns was low in both cases according to cross-validation results. Castro-Franco et al (2017) reported more consistent goodness of fit metrics for training and validation sets selected at random. Due to autocorrelation, observations held out at random closely resemble the training set, and validation metrics tend to be overestimated (Roberts et al, 2017).…”

Self-organizing map of soil properties in the context of hydrological modeling

“…Firstly, the watershed was segmented in three parts (low, medium and highlands) to establish for each the restriction values of soil depth for local minimum and maximum values [20]. Secondly, the land cover was used to identify the minimum soil depth understood as effective soil depth for root development, because of the sensitivity to this parameter affecting the hydrology dynamics and plant growth [30]. The covers considered were cropland, forest and bare soil; and the maximum root depth were assigned from the averages of crops and forest from literature references [31,32].…”

Hydrological modelling improvements in the assessemnt of water resources of agrarian subbasins in semiarid regions

“…Firstly, the watershed was segmented in three parts (low, medium and highlands) to establish for each the restriction values of soil depth for local minimum and maximum values (Rivas-Tabares et al, 2019b). Secondly, the land cover was used to identify the minimum soil depth understood as effective soil depth for root development, because of the sensitivity to this parameter affecting the hydrology dynamics and plant growth (Castro-Franco et al, 2017). The covers considered were cropland, forest and bare soil; and the maximum root depth were assigned from the averages of crops and forest from literature references (Canadell et al, 1996;Padilla and Pugnaire, 2007).…”