Browsing by Author "Somos-Valenzuela, Marcelo"
Now showing 1 - 3 of 3
Results Per Page
Sort Options
Item Early Estimation of Tomato Yield by Decision Tree Ensembles(2022) Lillo-Saavedra, Mario; Espinoza-Salgado, Alberto; García-Pedrero, Angel; Souto, Camilo; Holzapfel, Eduardo; Gonzalo-Martín, Consuelo; Somos-Valenzuela, Marcelo; Rivera Salazar, DiegoCrop yield forecasting allows farmers to make decisions in advance to improve farm management and logistics during and after harvest. In this sense, crop yield potential maps are an asset for farmers making decisions about farm management and planning. Although scientific efforts have been made to determine crop yields from in situ information and through remote sensing, most studies are limited to evaluating data from a single date just before harvest. This has a direct negative impact on the quality and predictability of these estimates, especially for logistics. This study proposes a methodology for the early prediction of tomato yield using decision tree ensembles, vegetation spectral indices, and shape factors from images captured by multispectral sensors on board an unmanned aerial vehicle (UAV) during different phenological stages of crop development. With the predictive model developed and based on the collection of training characteristics for 6 weeks before harvest, the tomato yield was estimated for a 0.4 ha plot, obtaining an error rate of 9.28%.Publication Earthquake-induced landslides coupled to fluvial incision in Andean Patagonia: inferring their effects on landscape at geological time scales(2022) Morales, Bastian; Lizama, Elizabet; Somos-Valenzuela, Marcelo; Rivera, Diego; Ningshen, ChenEarthquakes can deeply erode the mountainous landscape through co-seismic landslides, generating large amounts of sediment and debris that are then transported and distributed by rivers, controlling the landscape evolution. We can observe this influence in the Liqui˜ne Ofqui Fault System (LOFS), an active intra-arc fault system extending hundreds of kilometers through the Andes in Chilean Patagonia. For example, on April 21, 2007, a 6.2 Mw earthquake in the Ays´en Fjord triggered over 500 landslides with volumes reaching 12-20 Mm3. Although there is a well-defined seismic cycle, no study has focused on the effects of co-seismic landslides and sedimentary dynamics on the evolution of this mountainous landscape. In this research, we seek to improve the long-term understanding of the interaction between landslides and fluvial incisions in this segment of the Andes. For this reason, we implemented the Landlab-HyLands landscape evolution model (LEM), a hybrid landscape evolution model that allows modeling landslide activity coupled to fluvial incision. We consider the landslides that occurred during the 2007 earthquake as a precedent and simulate nine scenarios of ten seismic cycles over 21,000 years based on the 2100-year seismic cycle of the Holocene documented in this region. We further used multiple uplift rates, sediment erodibility, and m/n constant ratios associated with the current power law to assess this parameterization’s impact on the landscape. According to our results, landslides are a fundamental mechanism in the landscape’s evolution in this region. Deposits derived from landslides can create transitory landscape forms that can intervene in fluvial dynamics. According to our simulations, a significant part of the landslide sediment can remain on the slopes for thousands of years. We identified that parameterization considerably impacts the evolutionary response of the landscape in the evaluated time scale. Low m/n ratios can generate a different evolutionary response than other scenarios because the slopes are constantly driven towards their threshold angle, intensifying the interaction between landslides and fluvial incisions. Based on our analysis and considering the historical record of the Aysen Fjord, we can explain a critical primary control of the LOFS on landscape erosion and sediment production because of the surface seismic cycle. In our study, we demonstrate how the implementation of hybrid LEM can help to infer the contribution of sediments associated with large earthquakes and to improve the understanding of the role of landslides in the evolutionary history of Andean Patagonia. However, we stress that it is essential to advance in capturing erodibility and incision parameters of the current power law in the Andes and local geomechanical information. Finally, we believe the LEM can help to deepen the knowledge of these processes in other Andean basins exposed to these geomorphological processes.Item Ex Post Analysis of Water Supply Demand in an Agricultural Basin by Multi-Source Data Integration(2021) Lillo-Saavedra, Mario; Gavilán, Viviana; García-Pedrero, Angel; Gonzalo-Martín, Consuelo; Hoz, Felipe de la; Somos-Valenzuela, Marcelo; Rivera, DiegoIn this work, we present a new methodology integrating data from multiple sources, such as observations from the Landsat-8 (L8) and Sentinel-2 (S2) satellites, with information gathered in field campaigns and information derived from different public databases, in order to characterize the water demand of crops (potential and estimated) in a spatially and temporally distributed manner. This methodology is applied to a case study corresponding to the basin of the Longaví River, located in south-central Chile. Potential and estimated demands, aggregated at different spatio-temporal scales, are compared to the streamflow of the Longaví River, as well as extractions from the groundwater system. The results obtained allow us to conclude that the availability of spatio-temporal information on the water availability and demand pairing allows us to close the water gap—i.e., the difference between supply and demand—allowing for better management of water resources in a watershed