M. DJERBOUAI Salim

Soil erosion significantly impacts dam functionality by leading to reservoir siltation, reducing capacity, and heightening flood risks. This study aims to map soil erosion within a Geographic Information Systems (GIS) framework to estimate the siltation of the K'sob dam and compare these estimates with bathymetric observations. Focused on one of the Hodna basin’s sub-basins, the K'sob watershed (1477 km2), the assessment utilizes the Revised Universal Soil Loss Equation (RUSLE) integrated with GIS and remote sensing data to predict the spatial distribution of soil erosion. Remote sensing data were pivotal in updating land cover parameters critical for RUSLE, enhancing the precision of our erosion predictions. Our results indicate an average annual soil erosion rate of 7.83 t/ha, with variations ranging from 0 to 224 t/ha/year. With a typical relative error of about 13% in predictions, these figures confirm the robustness of our methodology. These insights are crucial for crafting mitigation strategies in areas facing high to extreme soil loss and will assist governmental agencies in prioritizing actions and formulating effective soil erosion management policies. Future studies should explore the integration of real-time data and advanced modeling techniques to further refine these predictions and expand their applicability in similar environmental assessments.

Monthly precipitation gaps filling using long short-term memory deep neural networks: Case of Soummam basin.

Recently, coupled Wavelet transform and Neural Networks models (WANN) were extensively used in hydrological drought forecasting, which is an important task in drought risk management. Wavelet transforms make forecasting model more accurate, by extracting information from several levels of resolution. The selection of an adequate mother wavelet and optimum decomposition level play an important role for successful implementation of wavelet neural network based hydrologic forecasting models.

The main objective of this research is to look into the effects of various discrete wavelet families and the level of decomposition on the performance of WANN drought forecasting models that are developed for forecast drought in the Algerois catchment for long lead time. The Standard Precipitation Index (SPI) is used as a drought measuring parameter at three-, six- and twelve-month scales. Suggested WANN models are tested using 39 discrete mother wavelets derived from five families including Haar, Daubechies, Symlets, Coiflets and the discrete approximation of Meyer. Drought is forecasted by the best model for various lead times varying from 1-month lead time to the maximum forecast lead time. The obtained results were evaluated using three performance criteria (NSE, RMSE and MAE).

The results show that WANN models with discrete approximation of Meyer have the best forecast performance. The maximum forecast lead times are 36-month for SPI-12, 18-month for SPI-6 and 7- month for the SPI-3. Drought forecasting for long lead times have significant values in drought risk and water resources management.

Recently, coupled Wavelet transform and Neural Networks models (WANN) were extensively used in hydrological drought forecasting, which is an important task in drought risk management. Wavelet transforms make forecasting model more accurate, by extracting information from several levels of resolution. The selection of an adequate mother wavelet and optimum decomposition level play an important role for successful implementation of wavelet neural network based hydrologic forecasting models.

The main objective of this research is to look into the effects of various discrete wavelet families and the level of decomposition on the performance of WANN drought forecasting models that are developed for forecast drought in the Algerois catchment for long lead time. The Standard Precipitation Index (SPI) is used as a drought measuring parameter at three-, six- and twelve-month scales. Suggested WANN models are tested using 39 discrete mother wavelets derived from five families including Haar, Daubechies, Symlets, Coiflets and the discrete approximation of Meyer. Drought is forecasted by the best model for various lead times varying from 1-month lead time to the maximum forecast lead time. The obtained results were evaluated using three performance criteria (NSE, RMSE and MAE).

The results show that WANN models with discrete approximation of Meyer have the best forecast performance. The maximum forecast lead times are 36-month for SPI-12, 18-month for SPI-6 and 7- month for the SPI-3. Drought forecasting for long lead times have significant values in drought risk and water resources management.

Abstract: This paper aims to identify the mineralisation origin and distinguish between the different classes of groundwater quality in several regions of the semiarid basin of Hodna in central Algeria. Many multivariate statistical techniques are applied to a dataset composed of 64 georeferenced individuals with 19 chemical variables. The obtained results from this principal component analysis show that the first five factors explain more than 78% of the groundwater quality variance. Other methods such as CA cluster analysis, CAH hierarchical cluster analysis and geochemical analysis using the Piper diagram are more appropriate to contemplate nodule-facies development and to distinguish clusters. The endorheic characteristic of the study basin consequents the basin centre named Chott El Hodna to be a salinity source. Bit by bit, salinity raises from North to South, from unsalted water to strongly salted water close to the Chott. The outcomes demonstrate that this groundwater is portrayed, the facies definite assessment outlines that the chloride, calcium and magnesium sulphate facies indicate 84% of cases, trailed by sodium sulphate facies with 14% and the rest (2%) is identified by the bicarbonate facies.

Keywords: GIS; groundwater quality; Hodna; hydrogeochemistry; multivariate statistical analysis; piper hydrochemical facies; principal component analysis; PCA; Algeria.

DOI: 10.1504/IJHST.2021.10040507

Recently, coupled Wavelet transform and Neural Networks models (WANN) were extensively used in hydrological drought forecasting, which is an important task in drought risk
management. Wavelet transforms make forecasting model more accurate, by extracting
information from several levels of resolution. The selection of an adequate mother wavelet
and optimum decomposition level play an important role for successful implementation of
wavelet neural network based hydrologic forecasting models.
The main objective of this research is to look into the effects of various discrete wavelet
families and the level of decomposition on the performance of WANN drought forecasting
models that are developed for forecast drought in the Algerois catchment for long lead
time. The Standard Precipitation Index (SPI) is used as a drought measuring parameter
at three-, six- and twelve-month scales. Suggested WANN models are tested using 39
discrete mother wavelets derived from five families including Haar, Daubechies, Symlets,
Coiflets and the discrete approximation of Meyer. Drought is forecasted by the best model
for various lead times varying from 1-month lead time to the maximum forecast lead time.
The obtained results were evaluated using three performance criteria (NSE, RMSE and
MAE).
The results show that WANN models with discrete approximation of Meyer have the
best forecast performance. The maximum forecast lead times are 36-month for SPI-12,
18-month for SPI-6 and 7- month for the SPI-3. Drought forecasting for long lead times
have significant values in drought risk and water resources management.
Keywords Algerois catchment · Drought · Forecasting · Neural networks · SPI ·
Wavelet transforms

This study aims to estimate the eroded and transported sediment yields from the The Hodna basin (26,000 km2) situated in central Algeria by two
approaches. In the first model, the data of the gauged subbasins are extrapolated to the ungauged areas based on the homogeneity of factors that influence the water
erosion-sediment transport process. In this approach, the specific eroded and transported sediment yield in the Hodna basin is estimated to be 425 t/km2
/yr. In an alternative approach, the eroded yield is estimated by mapping erosion using the (RUSLE) in a GIS environment. The obtained results show a high eroded sediment
yield of approximately 610 t/km2 /yr. The observed difference between the results of the two approaches can be explained by the amount of sediment that is eroded but is not transported by runoff. These two methods show high eroded and transported sediment yield values in the Hodna basin region; these high yields may seriously threaten the central flat zone with progressive deposition.

Abstract
In a semi-arid region of Maadher, central Hodna (Algeria), groundwater is the main source for agricultural and domestic
purposes. Anthropogenic activities and the presence of climate change’s efects have a signifcant impact on the region’s
groundwater quality. This study’s goals were to use water quality indices to evaluate the groundwater’s quality and its suitability for drinking and irrigation, as well as to identify contaminated wells using a geographic information system (GIS)
and the spatial interpolation techniques of ordinary kriging and inverse distance weighting (IDW). The results reveal that
all water samples exceeded the World Health Organization’s standards for nitrate ions and had alarming concentrations of
calcium, chlorine, and sulfate (WHO). According to Piper’s diagram, the groundwater hydrochemical facies is composed
of the elements sulfate–chloride-nitrate-calcium (SO42−-Cl—NO3−-Ca2+ water type). The majority of samples fall into the
poor water category, slightly more than 10% fall into the very poor water category, and less than 10% fall into the good
to the excellent quality category, per the water quality indices, which classify samples in a similar manner. According to
irrigation water indices, every sample is suitable for irrigation. Depending on the direction of groundwater fow, the spatial
distributions of Ca2+, Na+, Mg2+, SO42−, and Cl− show that their concentrations are high north of the area and relatively
low south of Maadher village (Fig. 3). Nitrate concentrations are high in the majority of samples, particularly those close
to the Bousaada wadi. In most samples, particularly those close to the Bousaada wadi, nitrate levels are high. Various water
quality models were described, and GIS spatial distribution maps were created using standard kriging and inverse distance
weighting (IDW) techniques through selected semi-variograms predicted against measurements. To determine the origin of
mineralization and the chemical processes that take place in the aquifer—which include the precipitation and dissolution of
dolomite, calcite, aragonite, gypsum, anhydrite, and halite—the groundwater saturation index was calculated.
Keywords Groundwater quality · Water quality indices · Kriging method · Inverse distance Weighting (IDW) method ·
Saturation index · Maadher

ABSTRACT

Due to the spatiotemporal variability of precipitation and the complexity of physical processes involved, missing precipitation data estimation remains as a significant problem. Algeria, like other countries in the world, is affected by this problem. In the present paper, Long Short-Term Memory (LSTM) deep neural Networks model was tested to estimate missing monthly precipitation data. The application was presented for the K’sob basin, Algeria. In the present paper, the optimal architecture of LSTM model was adjusted by trial-and-error-procedure. The LSTM model was compared with the most widely used classical methods including inverse distance weighting method (IDWM) and the coefficient of correlation weighting method (CCWM). Finally, it was concluded that the LSTM model performed better than the other methods.
Keywords: Hodna, K’sob basin, missing precipitation data, long short-term memory, CCWM, IDWM

ABSTRACT
Due to the spatiotemporal variability of precipitation and the complexity of physical processes involved, missing
precipitation data estimation remains as a significant problem. Algeria, like other countries in the world, is affected
by this problem. In the present paper, Long Short-Term Memory (LSTM) deep neural Networks model was tested
to estimate missing monthly precipitation data. The application was presented for the K’sob basin, Algeria. In the
present paper, the optimal architecture of LSTM model was adjusted by trial-and-error-procedure. The
LSTM model was compared with the most widely used classical methods including inverse distance weighting
method (IDWM) and the coefficient of correlation weighting method (CCWM). Finally, it was concluded that the
LSTM model performed better than the other methods.
Keywords: hodna, K’sob basin, missing precipitation data, long short-term memory, CCWM, IDWM.

Abstract
Soil degradation due to erosion by water is a serious environmental problem for the integrated management of basins, affecting the soil and water resources in Algerian region. Pluvial flood has been increasingly understood as a major threat that has
presented a significant risk for many watersheds worldwide, estimation of runoff and sediment yield is primarily required
for watershed development planning involving soil and water conservation measures, considering runoff is responsible for
sediment detachment and their transport during the erosion processes. In this context, the phenomenon reaches spectacular
values in many Algerian watersheds; in this case, it is very severe in flood period, many models have been developed and
their application has been spread. In this study, water flow discharge and suspended sediment concentration have been modelled using sediment rating curve approach, this model is the large discussed model; it is the best significant equation for the
majority of Algerian basins, this equation which has a power law form (i.e. C = aQb, where a and b are fitted parameters),
explains, more than 72% for the whole floods observed during 17 years in Wadi El Maleh watershed, the flood contribution
in annual suspended yield is variable; it can reach more than 92%, which is the case of flood; January 19, 1985, while, at
inter-annual scale, the percentage is 24% and 43% in total water and suspended sediment yields, respectively; for all studied floods, a good logarithmic correlation between sediment rating curve parameters is observed, this outcome can help to
extrapolate this model to other events.
Keywords Flood · Soil degradation · Sediment rating curve · Wadi El Maleh · Algeria

Abstract. In the practice, the precipitations records are linked to the problem of missing data caused by fault in the rain gaging station. In hydrology, estimating missing precipitation data is a crucial task due to the spatiotemporal variability of precipitations, also the complexity of physical processes involved.
We have done a comparative study between missing precipitations data estimation methods as next: classical methods: Inverse distance weighting method (IDWM) ,Correlation Coefficient Weighting Method (CCWM), principal component analysis (PCAM) ; Method based on genetic algorithms : fixed functional set genetic algorithm method (FFSGAM), as a target to judge, which methods are better to assess missing precipitation data. The application of these methods has been done using data of five rain gaging stations situated in the Macta watershed. We have tested the methods using the most recommended criterions of comparison. With the end we have noted that all the methods used, gave good results of estimate. And all FFSGAM models gave results more powerful than all the other methods.
Keywords: Missing precipitation data, Genetic algorithms, Correlation, Weighing methods, PCA.

Les enregistrements de précipitations sont généralement liés au problème des données manquantes dû aux défauts de fonctionnement dans les stations pluviométriques. L'évaluation des valeurs manquantes en hydrologie est un problème difficile du fait de la variabilité spatiotemporelle des précipitations et la complexité des processus physiques impliqués. Dans le but de juger les méthodes qui permettent de mieux estimer les données manquantes des précipitations dans le bassin versant de l’algérois au Nord de l’Algérie, nous avons mené une étude comparative entre les méthodes d’estimation: des méthodes classiques et des méthodes basées sur les algorithmes génétiques. Ces méthodes ont été testées sur deux pas de temps, journalier et mensuel, en utilisant les critères de comparaison les plus recommandés. Toutes les méthodes utilisées ont donné de très bons résultats d’estimation excepté la méthode IEWM. Les modèles basés sur les algorithmes génétiques ont donné des résultats plus performants que toutes les autres méthodes.


Mots clés :
Précipitations, algorithmes génétiques, corrélations, méthodes de pondération, ACP.

Abstract Drought forecasting is a major component of a drought preparedness and mitigation
plan. This paper focuses on an investigation of artificial neural networks (ANN) models for
drought forecasting in the algerois basin in Algeria in comparison with traditional stochastic
models (ARIMA and SARIMA models). A wavelet pre-processing of input data (wavelet
neural networks WANN) was used to improve the accuracy of ANN models for drought
forecasting. The standard precipitation index (SPI), at three time scales (SPI-3, SPI-6 and SPI-
12), was used as drought quantifying parameter for its multiple advantages. A number of
different ANN and WANN models for all SPI have been tested. Moreover, the performance of
WANN models was investigated using several mother wavelets including Haar wavelet (db1)
and 16 daubechies wavelets (dbn, n varying between 2 and 17). The forecast results of all
models were compared using three performance measures (NSE, RMSE and MAE). A
comparison has been done between observed data and predictions, the results of this study
indicate that the coupled wavelet neural network (WANN) models were the best models for
drought forecasting for all SPI time series and over lead times varying between 1 and 6 months.
The structure of the model was simplified in the WANN models, which makes them very
convenient and parsimonious. The final forecasting models can be utilized for drought early
warning.
Keywords Algerois catchment . ARIMA . Drought . Forecasting . Neural networks . SPI .
Wavelet transforms

Abstract : Drought forecasting is a major component of a drought preparedness and mitigation plan. This study compares linear stochastic models (ARIMA/SARIMA) and Feed-forward neural network (FFNN) models for drought forecasting in the Algerois catchment in Algeria, using standardized precipitation index (SPI) as a drought quantifying parameter. The results obtained from two models are presented in this paper.

Keywords : Drought; ARIMA;SARIMA; FFNN; Forecasting.

Drought forecasting is a major component of a drought preparedness and mitigation plan. This paper focuses on an investigation of artificial neural networks (ANN) models for drought forecasting in the algerois basin in Algeria in comparison with traditional stochastic models (ARIMA and SARIMA models). The standard precipitation index (SPI), at three time scales (SPI-3, SPI-6 and SPI-12), was used as drought quantifying parameter for its multiple advantages. A number of different ANN models for all SPI have been tested. The forecast results of all models were compared using three performance measures (NSE, RMSE and MAE). A comparison has been done between observed data and predictions, the results of this study are presented in this paper.

Keywords : Drought; ARIMA;SARIMA; ANN; Forecasting.

Drought forecasting is a major component of a drought preparedness and mitigation plan. In this study recursive multistep neural network (RMSNN) approach was used for drought forecasting in the Algerois catchment in Algeria, using standardized precipitation index (SPI) as a drought quantifying parameter.

Keywords : Drought; SPI; RMSNN; Forecasting.

Drought forecasting is a major component of a drought preparedness and mitigation plan. This study compares linear stochastic models known as multiplicative seasonal autoregressive integrated moving average (SARIMA) and recursive multistep neural network (RMSNN) for drought forecasting in the Algerois catchment in Algeria, using standardized precipitation index (SPI) as a drought quantifying parameter. The results obtained from two models are presented in this paper.