Christina Bogner

Characterising flow patterns in soils by feature extraction and multiple consensus clustering

Abstract The quality of surface water and groundwater is closely related to flow paths in the vadose zone. Therefore, dye tracer studies are often carried out to visualise flow patterns in soils. These experiments provide images of stained soil profiles and their evaluation demands knowledge in hydrology as well as in image analysis and statistics. The classical analysis consists of image classification in stained and non-stained parts and calculation of the dye coverage (i.e. the proportion of staining). The variation of this quantity with depth is interpreted to identify dominant flow types. While some feature extraction from images of dye-stained profiles is necessary, restricting the analysis to the dye coverage alone might miss important information. In our study we propose to use several index functions to extract different (ideally complementary) features. We associate each image row with a feature vector (i.e. a certain number of image function values) and use these features to cluster the image rows to identify similar image areas. Because images of stained profiles might have different reasonable clusterings, we calculate multiple consensus clusterings. Experts can explore these different solutions and base their interpretation of predominant flow type on quantitative (objective) criteria.

Is Ridge Cultivation Sustainable? A Case Study from the Haean Catchment, South Korea

Non-sustainable agricultural practices can alter the quality of soil and water. A sustainable soil management requires detailed understanding of how tillage affects soil quality, erosion, and leaching processes. Agricultural soils in the Haean catchment (South Korea) are susceptible to erosion by water during the monsoon. For years, erosion-induced losses have been compensated by spreading allochthonous sandy material on the fields. These anthropogenically modified soils are used for vegetable production, and crops are cultivated in ridges using plastic mulches. To evaluate whether the current practice of ridge cultivation is sustainable with regard to soil quality and soil and water conservation, we (i) analysed soil properties of topsoils and (ii) carried out dye tracer experiments. Our results show that the sandy topsoils have a very low soil organic matter content and a poor structure and lack soil burrowers. The artificial layering induced by spreading sandy material supported lateral downhill water flow. Ridge tillage and plastic mulching strongly increased surface runoff and soil erosion. We conclude that for this region a comprehensive management plan, which aims at long-term sustainable agriculture by protecting topsoils, increasing soil organic matter, and minimizing runoff and soil erosion, is mandatory for the future.

Image analysis for soil dye tracer infiltration studies

Flow processes in soils are closely related to groundwater quality often affected by human activities. Because hydrological models usually lack explanatory power, direct visualization of flow paths in dye tracer infiltration studies has become a standard tool in soil hydrology. These experiments provide images of dye-stained paths in soils and help evaluating the vulnerability or understanding the general hydrological functioning of a given site. Extracting relevant information demands expertise in hydrology as well as in image analysis and statistics. To our knowledge, no agreed and effective method to analyze large collections of such images exists in the soil hydrology community. In this paper we propose a general framework consisting of index functions and visual tools to support the expert in his/her evaluation of dye tracer infiltration images.

In‐situ prediction of soil organic carbon by vis–NIR spectroscopy: an efficient use of limited field data

Summary Visible–near-infrared diffuse reflectance spectroscopy (vis–NIR DRS) has been widely used to predict soil organic carbon (SOC) in the laboratory. Predictions made directly from soil spectra measured in situ under field conditions, however, remain challenging. This study addresses the issue of incorporating in-situ reflectance spectra efficiently into calibration data when a few field measurements only are available. We applied the synthetic minority oversampling technique (SMOTE) to generate new data with in-situ reflectance spectra from soil profiles. Subsequently, we combined existing spectral libraries with these new synthetic data to predict SOC by partial least squares regression (PLSR). We found that models with added synthetic spectra always outperformed models based on the spectral libraries alone and in most cases also those with added in-situ spectra only. We used the models to predict the distribution of SOC in soil profiles under five different land uses at Mount Kilimanjaro (Tanzania). Based on our results, we propose a framework for predicting SOC with a limited number of in-situ soil spectra. This framework could effectively reduce the costs of developing in-situ models for SOC at the local scale. Highlights We compare predictions of soil organic carbon from spectra of dried and sieved samples, field samples and calculated spectra. We use the synthetic minority oversampling technique (SMOTE) to calculate new soil spectra. Models with SMOTE outperform models with dried and field spectra in most cases. SMOTE can be used to reduce prediction errors when a few field data only are available for calibration.

Mapping Fractional Land Use and Land Cover in a Monsoon Region: The Effects of Data Processing Options

Existing global land use/land cover (LULC) raster maps have limited spatial and thematic resolution relative to the strong heterogeneity of agricultural landscapes. One promising approach to derive more informative maps is using fractional cover instead of hard classification. Here, we evaluate the effect of three key data processing options on the performance of random forest (RF) fractional cover models for moderate resolution imaging spectroradiometer (MODIS) data in a heterogeneous agricultural landscape in a monsoon region: 1) selection of spectral predictor sets [normalized difference vegetation index (NDVI), enhanced vegetation index (EVI), surface reflectance (SR), and all combined (Full)]; 2) time interval (8-day vs. 16-day); and 3) smoothing (no smoothing versus Savitzky-Golay (SG) filter). Model performance was assessed with spatially stratified rootmean-square error (RMSE), Spearmans rank correlation, and R2, per LULC type and averaged over all types. We found adequate performance of the best model (avg. ρ = 0.62) that used all predictors, 8-day interval and no smoothing. Among the different alternatives, the choice of predictors accounted for 36.3% of the variation, smoothing for 19.0%, and time interval for 17.9%. The intrinsic dimensionalities of the spectral predictors were investigated to complement the variable importance analyses. Although predicting LULC fractions for minor types remained difficult, our results suggest that existing satellite products can be a useful source of information about LULC at subpixel level provided the data-processing options are properly chosen.

Predicting with limited data — increasing the accuracy in vis-nir diffuse reflectance spectroscopy by smote

Diffuse reflectance spectroscopy is a powerful technique to predict soil properties. It can be used in situ to provide data inexpensively and rapidly compared to the standard laboratory measurements. Because most spectral data bases contain air-dried samples scanned in the laboratory, field spectra acquired in situ are either absent or rare in calibration data sets. However, when models are calibrated on air-dried spectra, prediction using field spectra are often inaccurate. We propose a framework to calibrate partial least squares models when field spectra are rare using synthetic minority oversampling technique (SMOTE). We calibrated a model to predict soil organic carbon content using air-dried spectra spiked with synthetic field spectra. The root mean-squared error of prediction decreased from 6.18 to 2.12 mg g−1 and R2 increased from −0.53 to 0.82 compared to the model calibrated on air-dried spectra only.

Classification of rare land cover types : Distinguishing annual and perennial crops in an agricultural catchment in South Korea

Many environmental data are inherently imbalanced, with some majority land use and land cover types dominating over rare ones. In cultivated ecosystems minority classes are often the target as they might indicate a beginning land use change. Most standard classifiers perform best on a balanced distribution of classes, and fail to detect minority classes. We used the synthetic minority oversampling technique (smote) with Random Forest to classify land cover classes in a small agricultural catchment in South Korea using modis time series. This area faces a major soil erosion problem and policy measures encourage farmers to replace annual by perennial crops to mitigate this issue. Our major goal was therefore to improve the classification performance on annual and perennial crops. We compared four different classification scenarios on original imbalanced and synthetically oversampled balanced data to quantify the effect of smote on classification performance. smote substantially increased the true positive rate of all oversampled minority classes. However, the performance on minor classes remained lower than on the majority class. We attribute this result to a class overlap already present in the original data set that is not resolved by smote. Our results show that resampling algorithms could help to derive more accurate land use and land cover maps from freely available data. These maps can be used to provide information on the distribution of land use classes in heterogeneous agricultural areas and could potentially benefit decision making.

Catchment Evapotranspiration and Runoff

The interplay between precipitation and evapotranspiration determines the input into the hydrological system of a catchment. Annual values of precipitation, evapotranspiration, and runoff measured at the catchment outlet for the 2002–2009 period were available. Annual precipitation clearly surmounted the sum of evapotranspiration and runoff. Part of the observed discrepancy might be due to the heterogeneity of precipitation and evapotranspiration within the catchment which has not been studied in sufficient detail. Annual evapotranspiration fluxes were remarkably constant during this period, whereas precipitation and runoff exhibited much larger interannual variability.

Dynamics of Water Flow in a Forest Soil: Visualization and Modelling

Soil water plays an important role in the terrestrial water and energy cycles. Its movement follows the gradient of the soil water potential and is most frequently described by the Richards equation. In this chapter, we focus on water fluxes in the vadose zone and model them with Water Heat and Nitrogen Simulation Model (WHNSIM) that solves the Richards equation numerically. We characterize the temporal dynamics of soil matric potentials measured at Coulissenhieb II and compare their complexity with modelled matric potential. Additionally, we summarize our previous studies on preferential flow—a common phenomenon in forest soils that cannot be modelled adequately by the Richards equation. The model WHNSIM reproduced the overall level of matric potentials in all depths. However, while it captured the complexity of the measurements in the upper soil, the matrix potentials in 90 cm depth were less complex indicating a more regular and damped signal. This result suggests that WHNSIM misses some important processes at least in the deeper soil. The soil water fluxes at Coulissenhieb II have a clear seasonal pattern with large fluxes occurring in spring during snow melt and small ones during dryer periods in summer. We could identify preferential flow in dye tracer experiments at the profile scale and attribute it mainly to macropore flow along root channels. Yet the identification and quantification of preferential pathways at the catchment scale remains challenging.