Browsing by Author "Acevedo Godoy, Sara Ester"

Now showing 1 - 4 of 4
  • Thumbnail Image
    Item
    A Combined and Extended Procedure for Measuring the Soil Water Retention and Hydraulic Conductivity Curves
    (2025) Contreras Torres, Cristina Pamela; Acevedo Godoy, Sara Ester; Avila Gorostiaga, Carlos Javier; Martinez, Sofia I.; Bonilla Melendez, Carlos Alberto
    Soil-specific properties like water retention and hydraulic conductivity are largely used in soil and environmental modelling and are typically obtained after laboratory analyses. So far, no single method is available to measure the entire suction range for water retention or hydraulic conductivity. Common methods for describing the soil water retention curve (SWRC) include simplified evaporation, pressure plates, neutron spectroscopy, and dewpoint. Regarding hydraulic conductivity, the techniques vary for the saturated or unsaturated condition, using tension disks and transient evaporation methods. In the search for a procedure to describe the entire water retention and hydraulic conductivity curves, the objective of this study was to illustrate the combination and use of a series of laboratory methods in eight different semi-hierarchical combinations to cover the whole suction range (0 <=$$ \le $$ pF <=$$ \le $$ 7). The data obtained from each combination was used to fit the van Genuchten-Mualem equation and compared using the RMSE and Akaike statistics. The main results show that using a combination of many methods for the water retention and hydraulic conductivity curves did not necessarily improve the curve fitting. However, adding data points at near saturation (pF close to 0) or from the driest part of the curve (pF >=$$ \ge $$ 4) improved the estimates on both curves. Specifically, for the clay soil, the RMSE for the hydraulic conductivity curve decreased from 0.0372 to 0.0369 cm/d when measurements from near saturation were added. For the sandy loam 2 soil, the RMSE for the water retention curve decreased from 0.039 to 0038 when including data from the driest part of the curve. Among all the soil-water-related parameters tested in this study, the estimates for the water retention content at the permanent wilting point (theta 1500 kPa) showed the largest difference among all the combinations of methods, up to 52%. In contrast, the difference in the water content at field capacity (theta 33 kPa) estimates was only 3%. This study provides an evaluation and insights to identify the best combination of methods when measuring or parametrizing the soil water retention and hydraulic conductivity curves.
  • Thumbnail Image
    Item
    Evaluation of machine learning algorithms for estimating soil water characteristics obtained by basic soil data in Central Chile
    (2023) Acevedo Godoy, Sara Ester; Bonilla Meléndez, Carlos Alberto; Pontificia Universidad Católica de Chile. Escuela de Ingeniería
    Las curvas de retención de agua en el suelo (SWRC) son mediciones hidrofísicas de alto costo y que requieren mucho tiempo, y que a menudo no son incluidas en los estudios de suelos ni en los análisis de rutina. Esta falta de mediciones hidrofísicas del suelo hace necesario el desarrollo de funciones de pedotransferencia (PTF), para determinados valores de succión (PTF puntuales) o para parámetros de forma de la curva (PTF paramétricas). Por lo tanto, la construcción de PTFs específicas de una región requiere conjuntos de datos detallados ricos en variabilidad, que son escasos. El objetivo de este estudio fue desarrollar un conjunto de datos hidrofísicos de suelos para los suelos chilenos, con el fin de complementar la falta de mediciones regionales de capacidad de campo (FC) y los parámetros de van Genuchten. La medición de la capacidad de campo (FC) es importante para comprender con precisión la dinámica hídrica del suelo. Por lo tanto, la FC debe obtenerse a partir de muestras de suelo inalteradas, para representar la condición natural del suelo. La reducida disponibilidad de datos de FC inalterada y la llegada de mediciones alternativas (métodos simplificados de evaporación y punto de rocío) han abierto muchas oportunidades para el desarrollo de funciones de pedotransferencia (PTFs) y la exploración de variables predictoras. Además, en este estudio se desarrolló un conjunto de datos hidrofísicos del suelo para comparar el ajuste VG unimodal con el bimodal en una amplia gama de propiedades del suelo. Basándose en este conjunto de datos, se evaluaron algoritmos de aprendizaje automático (redes neuronales artificiales, bosque aleatorio y Cubist) para la estimación de parámetros individuales de θs, θr, α1, n1, α2, n2 y w2 y FC en diferentes condiciones (disturbadas y no disturbadas). Nuestros resultados mostraron que los datos SWRC obtenidos utilizando el método de evaporación simplificada (SEM) junto con el método del punto de rocío (DP) conducían a un mejor ajuste cuando se utilizaba la parametrización VG bimodal, en la que la bimodal proporcionaba valores de θr más realistas que los obtenidos utilizando la parametrización VG unimodal. Se propuso y probó un conjunto de PTF para FC y parámetros bimodales. Con respecto al rango de succión SWRC, los conjuntos propuestos de PTF paramétricas tuvieron menos sesgo en el rango seco. Los resultados indican que el conjunto de datos hidrofísicos del suelo desarrollado y las PTF bimodales paramétricas derivadas pueden incluir la bimodalidad en la predicción de SWRC. Los resultados demostraron que si se tienen en cuenta las variables estructurales, se pueden desarrollar con éxito PTFs adecuadas para predecir la FC (tanto en muestras perturbadas como no perturbadas). Las PTFs desarrolladas en este estudio tienen métricas de precisión similares a las de estudios previos, pero utilizando menos predictores y siendo adecuados para predecir la FC en diferentes tipos de muestras de suelo. Por último, se evaluaron las propiedades físicas del suelo cerca de la superficie afectadas por la temperatura provocada por los incendios forestales. Los parámetros que describen el SWRC pueden obtenerse midiéndolos o derivándolos mediante funciones de pedotransferencia (PTF). Sin embargo, las PTF se han desarrollado utilizando datos de suelos agrícolas sin grandes eventos de calentamiento; por lo tanto, no es seguro que la estimación de los parámetros en suelos afectados por incendios sea fiable. El ampliamente utilizado PTF Rosetta se comparó con mediciones reales en dos condiciones: no afectado y afectado por el fuego. Las implicaciones para la modelización hidrológica se tradujeron en un menor contenido anual de agua y una mayor infiltración cuando se utilizaron las entradas de Rosetta en comparación con las entradas basadas en el SWRC medido.
  • No Thumbnail Available
    Item
    In-tandem methodology and recommendations for measuring soil hydraulic properties from saturation to dryness in laboratory conditions
    (wiley, 2025) Contreras Torres, Cristina Pamela; Acevedo Godoy, Sara Ester; Avila Gorostiaga, Carlos Javier; Bonilla Melendez, Carlos Alberto
    © 2025 Institute of Agrophysics, Polish Academy of Sciences.Laboratory determination of soil water retentions and hydraulic conductivity curves requires different techniques, depending on the number of data points needed. An important consideration in technique selection is soil sampling, particularly the distinction between disturbed and undisturbed samples, as laboratory methods can differ significantly for each kind of soil sample. This work addressed the increasing interest within soil physics in using methodologies that cover a wide range of water content and suction to determine soil hydraulic properties accurately. Achieving soil samples representative of physical processes is challenging due to the time required for measurements and the combination of techniques necessary to characterize soil water retention and hydraulic conductivity curves. A detailed workflow in-tandem for soil hydraulic curve measurements is presented, comprising more than 4 500 measurements to describe both curves from saturation to dryness. To enhance reproducibility and reduce uncertainty in laboratory measurement, this work offers ten best practice recommendations focused on sample preparation, instrument handling, and data management. This work aims to contribute to improved methodologies in soil physics laboratories and promote standardized practices. This work also contributes to advancing the reliability and accuracy of soil hydraulic properties assessment, thereby supporting improved research and application in soil science.
  • Thumbnail Image
    Item
    Testing the integral suspension pressure method for soil particle size analysis across a range of soil organic matter contents
    (2021) Acevedo Godoy, Sara Ester; Contreras Torres, Cristina Pamela; Ávila Gorostiaga, Carlos Javier; Bonilla Meléndez, Carlos Alberto; CEDEUS (Chile)
    Particle-size distribution is a critical part of soil description, which is commonly measured using pipette and hydrometer methods. However, a recently developed technique, called the integral suspension pressure method, allows for the measurement of continuous particle-size distribution based on Stokes' law. The objective of this study was to evaluate the applicability of the integral suspension pressure method for measuring particle-size distribution, as an alternative to the standard hydrometer procedure. The integral suspension pressure method was tested by using a soil dataset with a wide range of organic matter contents (0.22-12.0%). Forty-nine samples were analysed with a hydrometer after organic matter removal and the results were compared with those obtained using the integral suspension pressure method. Through comparing the integral suspension pressure and hydrometer measurements, root mean square error values of 8.9, 8.1, and 11.9% were observed for sand, silt, and clay, respectively. The clay fraction was underestimated throughout the entire range of measurements. Conversely, the silt content was overestimated over the whole range of measurements, especially in samples with more than 36% silt. When compared to the hydrometer method, integral suspension pressure integral suspension pressure exhibited a tendency to misclassify the soil texture of clay loam samples but was accurate for sandy loams.