A spatio-temporal hybrid neural network-Kriging model for groundwater level simulation

Tapoglou, Evdokia; Karatzas, George P.; Trichakis, Ioannis; Varouchakis, Emmanouil A.

doi:10.1016/j.jhydrol.2014.10.040

Cited by 89 publications

(26 citation statements)

References 57 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Despite the importance of groundwater as a component of the hydrological cycle and source of potable water, especially in arid and semiarid regions, there is a paucity of studies of long‐term groundwater fluctuations. One of the main reasons for this is that long‐term series observation data of groundwater depth are unavailable, and it is difficult to elucidate the rules of groundwater depth by using short‐term data (Dogrul et al, ; Jackson et al, ; Tapoglou et al, ). In this study, based on the significant correlation between groundwater depth and plant growth, we reconstructed groundwater depths during the past 66 years (1950–2015 CE) by using tree rings of P. euphratica in the lower reaches of the Tarim River (Figure b).…”

Section: Discussionmentioning

confidence: 99%

“…Moreover, the longest records of groundwater level histories measured by instruments are not more than 150 years in Europe and the United States (Liu, Gou, Zhang, & Chen, ). Consequently, it is necessary to reconstruct groundwater levels by utilizing a variety of methods to lengthen relatively short observational records (Dogrul, Kadir, Brush, & Chung, ; Herrmann, Baghdadi, & Blaschek, ; Jackson et al, ; Singh, Eldho, & Kumar, ; Tapoglou, Karatzas, Trichakis, & Varouchakis, ).…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Tree rings: A key ecological indicator for reconstruction of groundwater depth in the lower Tarim River, Northwest China

et al. 2019

View full text Add to dashboard Cite

Groundwater is becoming increasingly important in extremely arid areas due to climate change and human activities. Little is known, however, regarding the longterm dynamics of groundwater due to the lack of long-term historical records of groundwater depth data. In this study, we investigated the relationship between hydrometeorology and tree rings of Populus euphratica (P. euphratica) and the feasibility of reconstructing historical groundwater depths from these tree rings in the lower reaches of the Tarim River, an extremely arid area in China. Tree-ring growth of P. euphratica mainly reflected historical changes in regional water environments (including runoff and shallow groundwater), rather than climatic variables (including temperature, precipitation, relative humidity, and sunshine hours). Groundwater depth constituted the key environmental factor for directly controlling P. euphratica growth.Linear regression models using tree rings were developed to reconstruct historical changes of groundwater depths in the calendar year (January-December) and the growing season (April-August) from 1950 to 2015. Results tracked well with the main hydrological events in the study area. Our findings demonstrate that tree rings of P. euphratica are a key and effective ecological indicator for quantitatively reconstructing and elucidating historical changes of groundwater depth in the lower reaches of the Tarim River. Additionally, the suitable sampling range for establishing tree-ring chronologies of P. euphratica and reconstructing groundwater depth is 0-300 m from the river channel in the lower Tarim River. Taking into consideration the lagged effect between runoff of the Ecological Water Conveyance Project (EWCP), decrease in groundwater and tree growth, we recommend that the optimal implementation time of ecological water conveyance is between the previous September and November in the lower Tarim River. This application will effectively improve the groundwater environment in the current year and ensure the growth of riparian trees.

show abstract

Section: Discussionmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Tree rings: A key ecological indicator for reconstruction of groundwater depth in the lower Tarim River, Northwest China

et al. 2019

View full text Add to dashboard Cite

show abstract

“…All the spatial maps were prepared using ArcGIS 10.1 with Ordinary kriging (OK) model. This model is widely accepted and is employed in several pieces of research ( [41][42][43]. The Geochemist's Workbench was employed to understand the hydrochemical facies in the groundwater.…”

Section: Methodsmentioning

confidence: 99%

Coastal Aquifer Contamination and Geochemical Processes Evaluation in Tugela Catchment, South Africa—Geochemical and Statistical Approaches

Ntanganedzeni

Elumalai

Rajmohan

2018

Water

View full text Add to dashboard Cite

Abstract:Assessment of groundwater quality, contamination sources and geochemical processes in the coastal aquifer of Tugela Catchment, South Africa were carried out by the geochemical and statistical approach using major ion chemistry of 36 groundwater samples. Results suggest that the spatial distribution pattern of EC, TDS, Na, Mg, Cl and SO 4 are homogenous and elevated concentrations are observed in the wells in the coastal region and few wells near the Tugela River. Wells located far from the coast are enriched by Ca, HCO 3 and CO 3 . Durov diagrams, Gibbs plots, ionic ratios, chloro alkaline indices (CAI1 and CAI2) and correlation analysis imply that groundwater chemistry in the coastal aquifer of Tugela Catchment is regulated by the ion exchange, mineral dissolution, saline sources, and wastewater infiltration from domestic sewage; septic tank leakage and irrigation return flow. Principle component analysis also ensured the role of saline and anthropogenic sources and carbonates dissolution on water chemistry. Spatial distributions of factor score also justify the above predictions. Groundwater suitability assessment indicates that around 80% and 90% of wells exceeded the drinking water standards recommended by the WHO and South African drinking water standards (SAWQG), respectively. Based on SAR, RSC, PI, and MH classifications, most of the wells are suitable for irrigation in the study region. USSL classification suggests that groundwater is suitable for coarse-textured soils and salt-tolerant crops. The study recommends that a proper management plan is required to protect this coastal aquifer efficiently.

show abstract

“…It has been noted that ANFIS (Takagi and M. Sugeno, 1985;Jang, 1993Jang, , 1995Jang, , 1996Celikyilmaz and Turksen, 2009;Wang et al, 2009) exhibits better simulation performances than classical artificial neural networks (Nayak et al, 2004;El-Shafie et al, 2007;Firat, 2008;Pai et al, 2009;Wang et al, 2009;Maier et al, 2010). Moreover, ANFIS has already been successfully used to interpolate hydraulic head distribution (Lin and Chen, 2004;Kholghi and Hosseini, 2009;Flipo and Kurtulus, 2011;Kurtulus et al, 2011;Kurtulus and Flipo, 2012;Tapoglou et al, 2014).…”

Section: Introductionmentioning

confidence: 99%

Hydraulic head and groundwater <sup>111</sup>Cd content interpolations using empirical Bayesian kriging (EBK) and geo-adaptive neuro-fuzzy inference system (geo-ANFIS)

Sağır¹,

Kurtuluş²

2017

WSA

View full text Add to dashboard Cite

In this study, hydraulic head and 111 Cd interpolations based on the geo-adaptive neuro-fuzzy inference system (Geo-ANFIS) and empirical Bayesian kriging (EBK) were performed for the alluvium unit of Karabağlar Polje in Muğla, Turkey. Hydraulic head measurements and Cd content of groundwater. Both models were applied on the same case study: alluvium of Karabağlar Polje, which covers an area of 25 km 2 in Muğla basin, in the southwest of Turkey. The ANFIS method (called ANFIS XY ) uses two reduced centred pre-processed inputs, which are cartesian coordinates (XY). Geo-ANFIS is tested on a 100-random-data subset of 8 data among 42, with the remaining data used to train and validate the models. ANFIS XY and EBK were then used to interpolate hydraulic head and heavy metal distribution, on a 50 m 2 grid covering the study area for ANFIS XY , while a 100 m 2 grid was used for EBK. Both EBK-and ANFIS XY -simulated hydraulic head and 111 Cd distributions exhibit realistic patterns, with RMSE < 9 m and RMSE < 8 µg/L, respectively. In conclusion, EBK can be considered as a better interpolation method than ANFIS XY for both parameters.

show abstract

A spatio-temporal hybrid neural network-Kriging model for groundwater level simulation

Cited by 89 publications

References 57 publications

Tree rings: A key ecological indicator for reconstruction of groundwater depth in the lower Tarim River, Northwest China

Tree rings: A key ecological indicator for reconstruction of groundwater depth in the lower Tarim River, Northwest China

Coastal Aquifer Contamination and Geochemical Processes Evaluation in Tugela Catchment, South Africa—Geochemical and Statistical Approaches

Hydraulic head and groundwater <sup>111</sup>Cd content interpolations using empirical Bayesian kriging (EBK) and geo-adaptive neuro-fuzzy inference system (geo-ANFIS)

Contact Info

Product

Resources

About