Search

Despite advancements in sensor technology, monitoring nutrients in situ and in real-time is still challenging and expensive. Soft sensors, based on data-driven models, offer an alternative to direct nutrient measurements. However, the high demand for data required for their development poses logistical issues with data handling. To address this, the study aimed to determine the optimal subset of predictors and the sampling frequency for developing nutrient soft sensors using random forest. The study used water quality data at 15-min intervals from 2 automatic stations on the Main River, Germany, and included dissolved oxygen, temperature, conductivity, pH, streamflow, and cyclical time features as predictors. The optimal subset of predictors was identified using forward subset selection, and the models fitted with the optimal predictors produced R2 values above 0.95 for nitrate, orthophosphate, and ammonium for both stations. The study then trained the models on 40 sampling frequencies, ranging from monthly to 15-min intervals. The results showed that as the sampling frequency increased, the model's performance, measured by RMSE, improved. The optimal balance between sampling frequency and model performance was identified using a knee-point determination algorithm. The optimal sampling frequency for nitrate was 3.6 and 2.8 h for the 2 stations, respectively. For orthophosphate, it was 2.4 and 1.8 h. For ammonium, it was 2.2 h for 1 station. The study highlights the utility of surrogate models for monitoring nutrient levels and demonstrates that nutrient soft sensors can function with fewer predictors at lower frequencies without significantly decreasing performance. [ABSTRACT FROM AUTHOR]

Detailed knowledge of a soil's lime requirement (LR) is a prerequisite for a demand-based lime fertilization to achieve the optimum soil pH and thus sustainably increasing soil fertility and crop yields. LR can be directly determined by the base neutralizing capacity (BNC) obtained by soil-base titration. For a site-specific soil acidity management, detailed information on the within-field variation of BNC is required. However, soil-base titrations for BNC determination are too laborious to be extensively applied in routine soil testing. In contrast, visible and near-infrared spectroscopy (visNIRS) is a time and cost-effective alternative that can analyze several soil characteristics within a single spectrum. VisNIRS was tested in the laboratory on 170 air-dried and sieved soil samples of nine agricultural fields of a quaternary landscape in North-east Germany predicting the soil's BNC and the corresponding lime requirement (LRBNC) at a target pH of 6.5. Seven spectral pre-processing methods were tested including a new technique based on normalized differences (ND). Furthermore, six multivariate regression methods were conducted including a new method combining a forward stagewise subset selection algorithm with PLSR (FS-PLSR). The models were validated using an independent sample set. The best regression model for most target variables was FS-PLSR combined with the second Savitzky-Golay derivation as pre-processing method achieving R2s from 0.68 to 0.82. Finally, the performance of the direct prediction of LRBNC (R2 = 0.68) was compared with an indirect prediction that was calculated by the predicted BNC parameters. This resulted in slightly higher correlation coefficients for the indirect method with R2 = 0.75. [ABSTRACT FROM AUTHOR]