Browsing by Autor "Elvis Uscamayta-Ferrano"

Now showing 1 - 3 of 3

CHIRTS Gridded Air Temperature Downscaling Integrating MODIS Land Surface Temperature Estimates in Machine-Learning Models
(Multidisciplinary Digital Publishing Institute, 2025) Elvis Uscamayta-Ferrano; Frédéric Satgé; Ramiro Pillco Zolá; Henrique Llacer Roig; Diego Tola-Aguilar; María Eufemia Pérez-Flores; Lautaro Bustillos; Fara Pascale Rakotomandrindra; Zo Rabefitia; Simon Carrière
Due to its sensitivity to topographic and land use land cover features, air temperature (maximum, minimum, and mean—Tx, Tn, and Tmean) is extremely variable in space and time. The sparse and unevenly distributed meteorological stations observed across remote regions cannot monitor such variability. Freely available, gridded temperature datasets (T-datasets) are positioned as an opportunity to overcome this issue. Still, their coarse spatial resolution (i.e., ≥5 km) does not allow for the observation of air temperature variations on a fine spatial scale. In this context, a set of variables that have a close relationship with daily air temperature (MODIS maximum, minimum, and mean Land Surface Temperature—LSTx, LSTn, and LSTmean; MODIS NDVI; SRTM topographic features—elevation, slope, and aspect) are integrated in three regression machine-learning models (Random Forest—RF, eXtreme Gradient Boosting—XGB, Multiple Linear Regression—MLR) to propose a T-dataset estimates (Tx, Tn, and Tmean) spatial resolution downscaling framework. The approach consists of two main steps: firstly, the machine-learning models are trained at the native 5 km spatial resolution of the studied T-dataset (i.e., CHIRTS); secondly, the application of the trained machine-learning models at a 1 km spatial resolution to downscale CHIRTS from 5 km to 1 km. The results show that the method not only improves the spatial resolution of the CHIRTS dataset, but also its accuracy, with higher improvements for Tn than for Tx and Tmean. Among the considered models, RF performs the best, with an R2, RMSE, and MAE improvement of 2.6% (0%), 47.1% (6.1%), and 55.3% (7%) for Tn (Tx). These results will support air temperature monitoring and related extreme events such as heat and cold waves, which are of prime importance in the actual climate change context.
Development of Hourly Resolution Air Temperature Across Titicaca Lake on Auxiliary ERA5 Variables and Machine Learning-Based Gap-Filling
(Multidisciplinary Digital Publishing Institute, 2025) J. W. Sirpa-Poma; Juan Marcos Calle; Elvis Uscamayta-Ferrano; Jorge Molina‐Carpio; Frédéric Satgé; Osmar Cuentas Toledo; Ricardo Duran; Paula Pacheco Mollinedo; Rizuana Iqbal Hussain; Ramiro Pillco Zolá
This article presents an innovative procedure that combines advanced quality control (QC) methods with machine learning (ML) techniques to produce reliable, continuous, high-resolution meteorological data. The approach was applied to hourly air temperature records from six automatic weather stations located around Lake Titicaca in the Altiplano region of South America. The raw dataset contained time gaps, inconsistencies, and outliers. To address these, the QC stage employed Interquartile Range, Biweight, and Local Outlier Factor (LOF) statistics, resulting in a clean dataset. Two gap-filling methods were implemented: a spatial approach using time series from nearby stations and a temporal approach based on each station's time series and selected variables from the ERA5-Land reanalysis. Several ML models were also employed in this process: Random Forest (RF), Support Vector Machine (SVM), Stacking (STACK), and AdaBoost (ADA). Model performance was evaluated on a validation subset (30% of station data). The RF model achieved the best results, with R<sup>2</sup> values up to 0.9 and Root Mean Square Error (RMSE) below 1.5 °C. The spatial approach performed best when stations were strongly correlated, while the temporal approach was more suitable for locations with low inter-station correlation and high local variability. Overall, the procedure substantially improved data reliability and completeness, and it can be extended to other meteorological variables.
Machine-Learning Crop-Type Mapping Sensitivity to Feature Selection and Hyperparameter Tuning
(Multidisciplinary Digital Publishing Institute, 2026) Mayra Silvia Pérez-Flores; Frédéric Satgé; Paul Montesano; Renaud Hostache; Ramiro Pillco-Zolá; Diego Tola; Elvis Uscamayta-Ferrano; Lautaro Bustillos; Marie‐Paule Bonnet; Céline Duwig
To improve crop yields and incomes, farmers consistently adapt their practices to climate and market fluctuations, resulting in highly variable crop field distribution and coverage in space and time. As these dynamics illustrate farmers’ challenges, up-to-date crop-type mapping is essential for understanding farmers’ needs and supporting their adoption of sustainable practices. With global coverage and frequent temporal observations, remote sensing data are generally integrated into machine learning models to monitor crop dynamics. Unlike physical-based models that rely on straightforward use, implementing machine learning models requires extensive user interaction. In this context, this study assesses how sensitive the models’ outputs are to feature selection and hyperparameter tuning, as both processes rely on user judgment. To achieve this, Sentinel-1 (S1) and Sentinel-2 (S2) features are integrated into five distinct models (Random Forest (RF), Support Vector Machine (SVM), Light Gradient Boosting (LGB), Histogram-based Gradient Boosting (HGB), and Extreme Gradient Boosting (XGB)), considering several features selection (Variance Inflation Factor (VIF) and Sequential Feature Selector (SFS)) and hyperparameter tuning (Grid-Search) setup. Results show that the preprocess modeling feature selection (VIF) discards the features that the wrapped method (SFS) keeps, resulting in less reliable crop-type mapping. Additionally, hyperparameter tuning appears to be sensitive to the input features, and considering it after any feature selection improved the crop-type mapping. In this context a three-step nested modeling setup, including first hyperparameter tuning, followed by a wrapped feature selection (SFS) and additional hyperparameter tuning, leads to the most reliable model outputs. For the study region, LGB and XGB (SVM) are the most (least) suitable models for crop-type mapping, and model reliability improves when integrating S1 and S2 features rather than considering S1 or S2 alone. Finally, crop-type maps are derived across different regions and time periods to highlight the benefits of the proposed method for monitoring crop dynamics in space and time.