Harnessing Novel Data‐Driven Techniques for Precise Rainfall–Runoff Modeling

Journal of Flood Risk Management, Journal Year: 2025, Volume and Issue: 18(1)

Published: Feb. 17, 2025

ABSTRACT Rainfall and runoff are considered the main components of hydrological cycle, their forecasting is great significance in water resource management, particularly for reservoir operation. Developing an accurate model to capture dynamic connection between rainfall remains problematic challenging management due nonstationary characteristics hydrologic processes effects noise. In this study, data‐driven techniques, such as group method data handling (GMDH), extreme learning machine (ELM), two hybrids artificial neural network (ANN) with Cuckoo search algorithm (ANN + Cuckoo) genetic GA) were used rainfall–runoff relationship. For a comprehensive analysis, four scenarios examined based on different input combinations test select best scenario performance. The results indicated that performance ELM GMDH predicting was more than ANN GA. Although predicts higher accuracy, provides reliable simulating both low high values. models' can be ranked testing following order: GMDH, ELM, GA, CUKOO. root mean squared error (RMSE) recorded 56.7 69.7 m 3 /s models, respectively. These RMSE values highlight potential these models effectively addressing challenges associated complexity simulations. Moreover, demonstrate could simple, rapid, inexpensive approach timely prediction expected benefit management.

Language: Английский

---

Insight into the stable isotopic composition of glacial lakes in a tropical alpine ecosystem: Chirripó, Costa Rica

Hydrological Processes, Journal Year: 2018, Volume and Issue: 32(24), P. 3588 - 3603

Published: Sept. 22, 2018

Abstract Tropical high‐elevation lakes are considered sentinels of global climate change. This work characterizes the hydrological conditions tropical alpine glacial located in highlands Chirripó, Costa Rica, using a unique data set water stable isotopes (δ 2 H and δ 18 O) precipitation, stream water, lake between September 2015 July 2017. A combined dataset bathymetric, hydrometric, isotope collected 2016 2017 on Lake Ditkevi was used to calculate annual balance lake. Evaporation inflow ratios from three systems estimated linear resistance model, experimentally local evaporation line first lines region. The temporal isotopic variations O, d ‐excess, lc‐excess) confirm dry wet season evaporative for consistently average low ( E/I ) range 2.0 ± 0.8% 18.1 12.2%. Ditkevi's indicates steady‐state conditions, with an loss 650 mm/year (10.0 5.0% inflow), high‐water contribution catchment (90% residence time 0.53 0.27 years, scale (0.289 km yield or depth equivalent run‐off 278 mm/yr. These results provide novel information about losses lakes, which can serve as baseline future isotope‐based hydro‐climate research regions tropics elsewhere.

Language: Английский

---

Effect of land cover and hydro‐meteorological controls on soil water DOC concentrations in a high‐elevation tropical environment

Hydrological Processes, Journal Year: 2018, Volume and Issue: 32(17), P. 2624 - 2635

Published: July 6, 2018

Abstract Páramo soils store high amounts of organic carbon. However, the effects climate change and changes in land cover use (LC/LU) this high‐elevation tropical ecosystem may cause a decrease their carbon storage capacity. Therefore, better understanding factors influencing soils' export is urgently needed. To fill knowledge gap, we investigated differences dissolved (DOC) content soil water four LC/LU types (tussock grass, natural forest, pine plantations, pasture) controlling its variability Quinuas Ecohydrological Observatory south Ecuador. Weekly measurements DOC concentrations, meteorological variables, content, temperature from various depths slope positions were monitored within mineral horizons between October 2014 January 2017. These data used to generate regression trees random forest statistical models identify concentrations. From low depict highest concentrations followed by pasture, tussock forest. For all types, increase with decreasing moisture. Our results also show that most important predictor sampling depth Interestingly, atmospheric variables antecedent evapotranspiration precipitation conditions only little influence on during monitoring period. findings provide unique information can help improve management resources other peat dominated ecosystems elsewhere.

Language: Английский

---

A concerted research effort to advance the hydrological understanding of tropical páramos

Hydrological Processes, Journal Year: 2020, Volume and Issue: 34(24), P. 4609 - 4627

Published: Sept. 14, 2020

Abstract Páramos, a neotropical alpine grassland‐peatland biome of the northern Andes and Central America, play an essential role in regional global cycles water, carbon, nutrients. They act as water towers, delivering ecosystem services from high mountains down to Pacific, Caribbean, Amazon regions. Páramos are also widely recognized biodiversity climate change hot spots, yet they threatened by anthropogenic activities environmental changes. Despite their importance for security carbon storage, vulnerability human activities, only three decades ago, páramos were severely understudied. Increasing awareness need hydrological evidence guide sustainable management prompted action generating data filling long‐standing knowledge gaps. This has led remarkably successful increase scientific knowledge, induced strong interaction between scientific, policy, (local) communities. A combination well‐established innovative approaches been applied collection, processing, analysis. In this review, we provide short overview historical development research state hydrometeorology, flux dynamics, impacts, influence extreme events páramos. We then present emerging technologies hydrology resources discuss how converging science policy efforts have leveraged traditional new observational techniques generate base that can support conclude co‐evolution was able successfully cover different spatial temporal scales. Lastly, outline future directions showcase long‐term collection foster responsible conservation towers.

Language: Английский

---

End member and Bayesian mixing models consistently indicate near‐surface flowpath dominance in a pristine humid tropical rainforest

Hydrological Processes, Journal Year: 2021, Volume and Issue: 35(4)

Published: April 1, 2021

Abstract The impacts of forest conversion on runoff generation in the tropics have received much interest, but scientific progress is still hampered by challenging fieldwork conditions and limited knowledge about mechanisms. Here, we assessed generation, flow paths water source dynamics a pristine rainforest catchment Costa Rica using end member mixing analysis (EMMA) Bayesian model (MixSIAR). Geochemical tracer data collected over 4‐week field campaign were combined with tritium used to assess potential deeper groundwater pathways perennial stream. streamflow composition was best captured three end‐members, namely throughfall, shallow (5–15 cm) (15–50 soil water. We estimated end‐member contributions main stream two tributaries approaches found good agreement between results obtained from EMMA MixSIAR. system overwhelmingly dominated near‐surface sources, little evidence for older as tritium‐derived baseflow mean transit time 2.0 4.4 years. pathway (median 39% 49% based MixSIAR, respectively), followed (32% 31%) throughfall (25% 19%). had even greater relative (83% 74% tributary A 42% 63% B). Tributary B no detectable deep contribution, reflecting morphology hillslope (steeper slopes, shallower soils lower vegetation density compared A). Despite short sampling associated uncertainties, this study allowed thoroughly mechanisms humid tropical catchment. Our also provide first comparison increasingly models suggest that MixSIAR yield comparable estimates partitioning tropical, volcanic environment.

Language: Английский

---

Analysis of Stable Isotopes of Runoff Sources and their Implication in Hydropower Projects, Cases of Study of Manduriacu and Coca River Basins in Northern Ecuador

Research Square (Research Square), Journal Year: 2025, Volume and Issue: unknown

Published: May 8, 2025

Understanding the contributions of water sources to hydropower is crucial for characterizing hydrological processes in watersheds with limited data and resources. Specifically, these Ecuador are more complex, long-term monitoring limited; hence, estimating runoff challenging. In order unde0.rstand processes, stable isotopes δ¹⁸O δD were analyzed from monthly samples various tributaries draining northern Andean watersheds, encompassing elevations 500 5,897 meters, which contribute generation both projects. The sampling period was August 2017 September 2018. comparative analysis different headwater helps characterize runoff, contributing projects originating glaciers Paramo system. Malo Mashpi Rivers, located Amazon Pacific drainage basins, respectively, do not receive high-elevation water; instead, they represent local sources. These two rivers join downstream included process purposes, despite influencing generation. significant role altitude isotope composition rainfall-runoff waters indicates that Manduriacu predominantly receives paramo systems. Lower elevation govern Coca-Codo Sinclair Hydropower intake project on eastern side. From a management perspective, this research found 95% total energy produced systems came rainforest rainfall during period.

Language: Английский

---

Changes in Soil Hydro-Physical Properties and SOM Due to Pine Afforestation and Grazing in Andean Environments Cannot Be Generalized

Forests, Journal Year: 2018, Volume and Issue: 10(1), P. 17 - 17

Published: Dec. 29, 2018

Andean ecosystems provide important ecosystem services including streamflow regulation and carbon sequestration, that are controlled by the water retention properties of soils. Even though these soils have been historically altered pine afforestation grazing, little research has dedicated to assessment such impacts at local or regional scales. To partially fill this knowledge gap, we present an evaluation plantations grazing on soil hydro-physical organic matter (SOM) high montane forests páramo in southern Ecuador, elevations varying between 2705 3766 m a.s.l. In total, seven study sites were selected each one was parceled into undisturbed plots with plantation grazing. Soil characterized two depths, 0–10 10–25 cm, differences parameters disturbed analyzed versus factors as type, sampling depth, elevation, past/present land management. The main affected use change saturated hydraulic conductivity (Ksat), capacity (pF 0 2.52), SOM. dependent characteristics, previous use, while primarily depth management (grazing intensity tilling activities). site-specific nature found relations suggests extension findings response changes is risky; therefore, future evaluations impact should take consideration responses can be site specific.

Language: Английский

---

Carbon Dioxide (CO₂) Fluxes From Terrestrial and Aquatic Environments in a High‐Altitude Tropical Catchment

Journal of Geophysical Research Biogeosciences, Journal Year: 2020, Volume and Issue: 125(8)

Published: July 28, 2020

Abstract High‐altitude tropical grasslands, known as “páramos,” are characterized by high solar radiation, precipitation, and low temperature. They also exhibit some of the highest ecosystem carbon stocks per unit area on Earth. Recent observations have shown that páramos may be a net source CO 2 to atmosphere result climate change; however, little is about this excess in these mountainous environments or which landscape components contribute most . We evaluated spatial temporal variability surface fluxes from adjacent terrestrial aquatic high‐altitude catchment Ecuador, based suite field measurements performed during wet season. Our findings revealed importance hydrologic dynamics regulating magnitude likely fate dissolved stream. While headwater catchments disproportionately larger amounts than their downstream counterparts, our study highlights heterogeneity within between elements complex topography. evasion stream surfaces was up an order greater soil efflux environment. Stream flux appeared transport limited (i.e., controlled flow characteristics, turbulent flow, water velocity) upper reaches stream, availability) lower A 4‐m waterfall along channel accounted for 35% total observed 250‐m reach. These represent first step understanding cycling at interface ecosystems high‐altitude, tropical, catchments.

Language: Английский

---

Soil–vegetation–water interactions controlling solute flow and chemical weathering in volcanic ash soils of the high Andes

Hydrology and earth system sciences, Journal Year: 2023, Volume and Issue: 27(7), P. 1507 - 1529

Published: April 11, 2023

Abstract. Vegetation plays a key role in the hydrological and biogeochemical cycles. It can influence soil water fluxes transport, which are critical for chemical weathering development. In this study, we investigated balance solute two profiles with different vegetation types (cushion-forming plants vs. tussock grasses) high Ecuadorian Andes by measuring content, flux, concentrations modeling hydrology. We also analyzed of weathering. The on is restricted to A horizon. Evapotranspiration 1.7 times higher deep drainage 3 lower under cushion-forming than grass. Likewise, cushions transmit about 2-fold less from horizons. This attributed retention saturated hydraulic conductivity associated shallower coarser root system. Under plants, dissolved organic carbon (DOC) metals (Al, Fe) mobilized Solute that be related plant nutrient uptake (Mg, Ca, K) decline depth, as expected biocycling nutrients. Dissolved silica bicarbonate minimally influenced represent largest contributions fluxes. Soil constant depth below grasses but declining plants. difference mainly Our findings reveal modify properties uppermost horizon, altering balance, fluxes, throughout profile.

Language: Английский

---

Wavelet analyses of neural networks based river discharge decomposition

Hydrological Processes, Journal Year: 2020, Volume and Issue: unknown

Published: Feb. 13, 2020

The problem of discharge forecasting using precipitation as input is still very active in Hydrology, and has a plethora approaches to its solution. But, when the objective simulate values without considering phenomenology behind processes involved, Artificial Neural Networks, ANN give good results. However, question how black box internally solve this remains open. In research, classical rainfall-runoff approached that total sum components hydrological system, which from perspective translated three signals related fast, middle slow flow. Thus, present study two aims (a) time-frequency representation by an hydrologic model (b) capabilities additively decompose river discharge. This adds knowledge open physical interpretability black-box models, limited. results show adequately simulated time frequency domain, although less power spectrum evident during rainy seasons model, due fast flow underestimation. wavelet represents well slow, system with transit times 256, 12–64 2–12 days, respectively. Interestingly, these are remarkably similar those soil water reservoirs studied small headwater catchment tropical Andes. result needs further research because it opens possibility determining MMT on fraction cost isotopic based methods. cross-power indicates error more misrepresentation components, despite limitations recharge period component. With respect individual flows neurons were uncapable individually represent such flows. combination pairs resemble dynamics spectral content aforementioned signals. These some evidence signal processing techniques may be used infer information about functioning basin.

Language: Английский

---