The research team selected the Kitchener Drain estuary as their proving ground, and for good reason. This drainage channel sits between the Mediterranean and the Nile Delta, one of the most vulnerable coastal regions on the planet, where natural erosion pressures collide with pollution, agricultural runoff and human modification of the coast. Sediment dynamics in such an environment are notoriously difficult to capture: waves reshape the seabed, currents redistribute sand, and drain outflows disturb the balance in ways that traditional models struggle to resolve quickly enough for practical decision-making.

The framework begins with physics. The researchers simulated hydrodynamic and wave interactions using the Coastal Modelling System, or CMS, a software package developed for coastal processes applications that has been verified and validated for sediment transport and morphology change. CMS computes how water moves through the estuary, how waves propagate and break, and how those forces mobilize and deposit sediment. These simulations generate a rich numerical picture of the system, but running them is computationally expensive, and tuning their parameters by hand is a slow, expert-driven process that can take weeks for a single scenario.

That is where the artificial intelligence enters. The team applied Bayesian optimization, a statistical technique for efficiently searching complex parameter spaces, to refine the hydrodynamic model’s settings automatically. Instead of an engineer laboriously adjusting friction coefficients and transport formulas one at a time, the optimization algorithm learns from each simulation which parameter combinations are most promising and concentrates its search accordingly. This step matters because poorly calibrated physics models produce garbage inputs, and no amount of downstream machine learning can rescue a badly calibrated foundation.

With the simulations tuned, the researchers trained five machine learning models to predict sediment dynamics, including Decision Trees, Random Forests and Gradient Boosting. The dataset combined numerical outputs from the simulations with experimental observations, giving the algorithms both the structured physics of the coastal system and real-world grounding. The goal was to build fast surrogate models that could reproduce the expensive hydrodynamic simulations in a fraction of the time, allowing managers to test many what-if scenarios instead of just a handful.

The results delivered a surprise for anyone who assumes bigger models are always better. The humble Decision Tree achieved the best overall predictive performance, with root mean square error values of 0.3282 plus or minus 0.0180 meters under 5-fold cross-validation and 0.3243 plus or minus 0.0230 meters under 10-fold cross-validation. Random Forests performed nearly identically, posting RMSE values of 0.3286 plus or minus 0.0202 meters and 0.3248 plus or minus 0.0241 meters for the same validation schemes. In practical terms, the models predict morphological change with errors on the order of a third of a meter, a precision that is genuinely useful for planning coastal interventions.

Cross-validation, the technique behind those error estimates, deserves a word of explanation because it is the difference between a trustworthy result and an inflated one. By splitting the data into five or ten partitions and repeatedly training on all but one partition before testing on the held-out set, the researchers ensured their error figures reflect performance on unseen data rather than memorization. The close agreement between the 5-fold and 10-fold results suggests the models are stable and not overfitting, a persistent danger when machine learning is applied to geophysical datasets of modest size.

Interpretability was treated as a first-class requirement rather than an afterthought. The team performed a feature importance analysis to identify which parameters drive the predictions, giving coastal engineers insight into why the model forecasts erosion or deposition at a given location. This matters enormously for adoption: a black-box prediction that a shoreline will retreat is hard to act on, but knowing that wave height, current velocity or drain discharge dominates the forecast tells engineers which lever to pull. The approach echoes a broader movement in the geosciences toward explainable machine learning, where the goal is not just accuracy but understanding.

Perhaps the most immediately practical contribution is a Graphical User Interface that wraps the trained models, enabling rapid and interactive prediction of morphological changes from user-provided hydrodynamic inputs. A coastal manager with no programming background can enter local conditions and receive near-instant forecasts of how the seabed and shoreline will respond. This kind of decision-support tooling is what transforms an academic exercise into something that can inform real choices about where to place defenses, how to manage dredging, and which stretches of coast deserve urgent protection.

The stakes could hardly be higher. The Nile Delta hosts millions of people, fertile agricultural land and critical infrastructure within a few meters of sea level, and studies of the Egyptian coast have documented persistent shoreline retreat, sea level rise impacts and sediment starvation following the damming of the Nile. The Kitchener Drain outlet itself has been the subject of earlier hydro-morphological investigations by overlapping research groups, and this new framework builds on that foundation by adding predictive speed and machine learning rigor. If the approach generalizes to other vulnerable estuaries, it could become a template for coastal forecasting in data-scarce regions worldwide, where running full physics simulations for every planning question is simply not feasible.

There are, of course, limits to keep in mind. The models were trained on data from a single estuary, and sediment transport is famously site-specific, so transferring the framework elsewhere would require recalibration and retraining with local hydrodynamic simulations. The authors also note that no new field datasets were generated for the study, meaning the machine learning component leans heavily on the fidelity of the CMS simulations and existing observations. Still, the architecture is deliberately modular: better physics, better data and better algorithms can each be swapped in as they improve. What the study demonstrates is a workflow, not just a result, and that workflow, physics simulation, Bayesian optimization, ensemble machine learning, interpretability analysis and GUI deployment, is one that other coastal engineering teams can adopt today. As climate change accelerates erosion on deltas from the Mediterranean to the Mekong, tools that turn weeks of computation into seconds of interaction may prove to be exactly the lifeline vulnerable coastlines need.

Subject of Research: Hybrid hydrodynamic simulation and machine learning for predicting coastal erosion and sedimentation at the Kitchener Drain estuary on the Nile Delta

Article Title: A hybrid hydrodynamic and machine learning framework for coastal erosion and sedimentation prediction

Article References: Elshaarawy, M. K., Nassar, K., & Mansour, N. (2026). A hybrid hydrodynamic and machine learning framework for coastal erosion and sedimentation prediction. Earth Science Informatics, 19(12), Article 216. https://doi.org/10.1007/s12145-026-02274-9

Image Credits: AI Generated

DOI: 10.1007/s12145-026-02274-9

Keywords: coastal erosion, machine learning, sediment transport, Nile Delta, Coastal Modelling System, Bayesian optimization, Decision Trees, Random Forests, hydrodynamic modeling, shoreline change, Kitchener Drain, sea level rise

The study, published in Environmental Monitoring and Assessment by Hesham M. El-Asmar and Mahmoud Sh. Felfla of Damietta University and Hussein M. Rashad of Egypt’s Ministry of Environment, comes at a pivotal moment for the lagoon. Between 2017 and 2022, Manzala underwent extensive restoration and dredging, reshaping its channels and circulation patterns. Managers suddenly needed up-to-date, high-resolution depth maps to feed hydrodynamic models, track sediment movement, and plan further interventions. Traditional surveying over such a large, shallow water body was impractical, so the researchers turned the problem over to algorithms.

The core idea behind satellite-derived bathymetry is elegant: light penetrates clear water, and the deeper the bottom, the dimmer and more color-shifted the reflected signal becomes when it finally reaches a sensor in orbit. Blue and green bands carry most of the depth information, while red light is absorbed quickly. But Manzala is anything but clear. Turbidity, suspended sediments, algal activity, and bottom vegetation all scramble the optical signal, which is precisely why simple empirical formulas developed for tropical coral reefs tend to collapse in such environments. Machine learning, the researchers reasoned, could learn the messy, nonlinear relationships between spectral reflectance and actual depth if it was given enough trustworthy ground truth.

That ground truth came from extensive echo-sounder measurements collected across the lagoon, which were paired with reflectance values extracted from Landsat 8 multispectral imagery. Seven machine-learning algorithms were then pitted against one another, spanning tree-based ensembles and boosting-based regressors. The evaluation was deliberately thorough: rather than relying on a single correlation coefficient, the team examined regression metrics, depth-class performance, uncertainty estimation, and, crucially, an interpretability analysis using SHAP values, a technique borrowed from explainable artificial intelligence that reveals how much each input feature contributes to each prediction.

The results were striking. Tree-based ensembles consistently outperformed the boosting-based regressors, with Decision Tree and Extra Trees models achieving the highest test-set accuracy under the adopted validation design, reaching a correlation coefficient of 0.996, an R-squared of 0.992, and a root-mean-square error of just 0.073 meters. In practical terms, the satellite-based predictions deviated from the echo-sounder measurements by less than eight centimeters on average, a level of precision that would have seemed implausible for a turbid lagoon only a few years ago.

Yet the researchers were careful not to crown a winner on accuracy alone. Two models that score identically on a test set can behave very differently when deployed across a full lagoon, particularly in areas where the optical conditions differ from the training data. When the team quantified prediction uncertainty, Extra Trees emerged as the more dependable choice, with uncertainty of roughly 0.09 meters, substantially lower than its equally accurate competitor. That distinction matters for operational mapping, where a map that is confidently wrong is far more dangerous than one whose errors are known and bounded. For the purposes of managing a living lagoon, the Extra Trees model was deemed the most reliable.

Perhaps the most scientifically interesting finding came from the SHAP analysis, which identified the near-infrared band as the single most influential predictor. On the surface, this is counterintuitive: near-infrared light is absorbed within the first few tens of centimeters of water, so it cannot directly sense the bottom of a lagoon several meters deep. The authors interpret its dominance as evidence that the model is exploiting surface and turbidity-related effects rather than true depth sensing, effectively using the near-infrared signal as a proxy for the optical state of the water column. This is a valuable caution for the rapidly growing field of satellite-derived bathymetry: a model can achieve spectacular statistics while leaning on correlations that may not generalize to other water bodies or seasons. The transparency offered by SHAP-based interpretation allows practitioners to spot such behavior before it leads them astray.

The study also highlighted the value of feature engineering rooted in physics. Logarithmic band ratios, a transformation descended from classical empirical bathymetry methods dating back to Lyzenga’s foundational work in the late 1970s and Stumpf’s ratio-based algorithms of the early 2000s, improved model robustness by capturing nonlinear spectral relationships while partially canceling out variable bottom reflectance. In other words, the best results came not from throwing raw pixels at a black-box algorithm, but from combining decades of optical oceanography with modern ensemble learning, then interrogating the result with explainability tools.

The final bathymetric maps resolved depth gradients spanning roughly negative 0.6 to negative 4.0 meters across the lagoon, capturing the shallow fringes, dredged channels, and deeper basins that govern how water moves through the system. Such maps are not merely cartographic trophies. They are the geometric foundation for hydrodynamic models that simulate circulation and flushing, for sediment transport studies that predict where dredging will need to be repeated, and for ecological assessments of habitats that depend on specific depth and light regimes. For a lagoon that supplies much of Egypt’s fish catch and buffers the densely populated Nile Delta coast, that information carries real economic and environmental weight.

Beyond Manzala, the framework’s significance lies in its scalability and cost profile. Landsat data are free, the algorithms are open and interpretable, and the only expensive ingredient, the echo-sounder calibration campaign, can be far smaller than a full survey because the machine-learning model interpolates between calibration points. The authors position the approach as a transparent, efficient alternative to exhaustive field programs, one that can be repeated as conditions change, whether through further restoration work, sedimentation, or sea-level-driven alterations to the Delta’s fragile coastal wetlands. As turbid shallow waters around the world face mounting pressure from climate change and development, the Manzala study offers a template: pair trusted field measurements with ensemble machine learning, demand uncertainty estimates alongside accuracy statistics, and never trust a model you cannot explain.

Subject of Research: Satellite-derived bathymetry mapping of a turbid coastal lagoon using comparative machine-learning models

Article Title: Comparative machine-learning retrieval of high-precision bathymetry from multispectral remote sensing: a case study of Manzala Lagoon, Egypt

Article References: El-Asmar, H. M., Felfla, M. S., & Rashad, H. M. (2026). Comparative machine-learning retrieval of high-precision bathymetry from multispectral remote sensing: a case study of Manzala Lagoon, Egypt. Environmental Monitoring and Assessment, 198(10), Article 1049. https://doi.org/10.1007/s10661-026-15878-x

Image Credits: AI Generated

DOI: 10.1007/s10661-026-15878-x

Keywords: satellite-derived bathymetry, machine learning, Landsat 8, Manzala Lagoon, Nile Delta, remote sensing, SHAP interpretability, Extra Trees, coastal lagoon management, echo-sounder calibration, turbid waters, Environmental Monitoring and Assessment

The research combines three complementary techniques, each filling in a different layer of the story. Sediment coring provided vertical slices through the landscape, revealing the sequence of natural and human-made deposits beneath the site. Electrical resistivity tomography and direct-current resistivity soundings mapped subsurface structures horizontally, tracing buried features without breaking the surface. Finally, phytolith analysis, the study of microscopic silica bodies formed within plant tissues, allowed the researchers to reconstruct which plants grew around the city and how vegetation changed as urbanization took hold. Together, these methods produced a picture of Bubastis that no single approach could deliver: a city whose growth, layout, and even its religious geography were dictated by the shape of the land beneath it.

The most consequential discovery concerns a set of Pleistocene sand mounds known as Geziras. These are relict landforms, remnants of ancient sandy deposits that rise subtly above the surrounding floodplain and, crucially, stay dry during the Nile’s annual inundation. Garbe’s surveys identified several of these mounds across the Bubastis area, and their positions are anything but random. The Temple of Pepi I in the western part of the city stands on one. The palace and the cemetery area in the north occupy others. Residential zones in eastern Bubastis cluster on still more. Even a lost monument may now have a location: an additional mound in the southeastern part of the site can plausibly be linked to the Temple of Hermes, a structure known from ancient writings whose whereabouts had been uncertain. The new evidence supports those textual clues about where the temple once stood.

This pattern reveals a deliberate urban logic. Rather than spreading indiscriminately across the floodplain, the builders of Bubastis selected flood-safe Geziras for their most important and most permanent structures, while leaving the low-lying ground between them for uses that could tolerate, or even benefit from, seasonal flooding. The geomorphology of the Delta was not an obstacle to be overcome but a framework to be exploited. In a landscape where the Nile’s branches shifted course and flood levels fluctuated over the centuries, choosing the right piece of ground was a matter of survival, and the people of Bubastis appear to have understood that from the city’s earliest days.

The coring program also quantified just how intensively different parts of the city were used. In settlement area S1, the researchers documented anthropogenic deposits reaching a thickness of up to 9.5 meters, an extraordinary accumulation of debris, building material, and occupation layers that identifies S1 as the primary living quarter of the city. The thickest deposits date particularly to the Ptolemaic period, from about 332 to 30 BCE, and the Roman period, from about 30 BCE to 200 CE, showing that these late phases of Bubastis’s history were also among its most densely built. By contrast, the surrounding low-lying floodplain areas, especially area S2, show only sparse settlement evidence. These marginal zones were likely given over to agriculture, and they had a distinct advantage for it: a nearby Nile branch, flowing to the north and northeast of the city, regularly deposited nutrient-rich sediments across the floodplain, keeping the fields naturally fertilized.

The phytolith record adds a biological dimension to this reconstruction, tracking vegetation from the end of the last ice age through the city’s decline. During the transition from the Late Pleistocene to the Early Holocene, long before Bubastis existed, the landscape was dominated by grasses, shrubs, and trees, a natural Delta environment shaped by river processes rather than human hands. That changed with the arrival of settlement. From the Old Kingdom onward, spanning roughly 2850 to 2180 BCE, the onset of occupation coincided with a marked increase in cereal-type phytoliths. Crucially, these cereal signatures appear in the nutrient-rich floodplain sediments, providing the earliest evidence for localized cereal cultivation at Bubastis during the Early Old Kingdom. The fields, in other words, were already feeding the city in its formative centuries.

Inside the settlement itself, the phytolith data tell a different story. Elevated proportions of cereal-type phytoliths in anthropogenic deposits across the urban area, and particularly in the main residential zone S1, point to intensive cereal processing and storage taking place within the city walls. Grain was not only grown nearby but handled, threshed, milled, and stockpiled in the residential quarters, a pattern that speaks to the scale of Bubastis’s population and the logistical sophistication required to sustain it. The city was, in a very real sense, an agricultural machine as much as a religious and administrative center, with its food economy woven directly into the fabric of its neighborhoods.

Over time, that arrangement shifted. From the Late Period, beginning around 664 BCE, through Roman times, cultivation practices increasingly moved toward the settlement periphery. As Bubastis expanded and its builders consumed more land, the natural vegetation around the city diminished, and farming was pushed outward to the edges of the urban footprint. This trajectory, from cultivation embedded within and immediately around the early settlement to agriculture relegated to the periphery of a sprawling city, mirrors patterns familiar from urban growth in other ancient and modern contexts, but here it is documented with unusual precision through a combination of sediment chemistry, landform mapping, and plant microfossils.

The broader significance of the work lies in what it says about human adaptability in unstable environments. The Holocene Nile Delta was shaped by a constant interplay of river avulsions, sedimentation, and seasonal flooding, conditions that created fertile opportunities for early civilization while simultaneously driving settlement instability and frequent relocation. Bubastis’s longevity was not an accident of geography but the product of choices: siting temples and palaces on high ground, farming the fertile lows, and adjusting land use as the city and its environment co-evolved. Different geomorphological settings were deliberately selected for specific functions and phases of occupation, a dynamic interplay between nature and human decision-making that unfolded over three thousand years.

The study also refines interpretations of Bubastis’s emergence as a center of religious and administrative power, a status reflected in its extensive temple and palace areas, including the famous Temple of Bastet with its man-made sacred canals identified in earlier geoarchaeological work. Until now, knowledge of the site’s environmental conditions rested on a handful of historical texts and preliminary studies. By delivering the first area-wide reconstruction, the new research demonstrates how strongly the development of an ancient Egyptian urban center depended on regional environmental conditions, and it showcases the value of geoarchaeological methods for reconstructing human-environment interactions across the Delta, where so many settlements rose and fell with the shifting waters of the Nile. For Bubastis, the ground beneath the city was never just a foundation. It was a partner in the city’s survival, and reading that partnership from the sediments is what finally explains how one Delta town outlasted them all.

Subject of Research: Holocene landscape evolution and long-term settlement development at the ancient Egyptian city of Bubastis in the southeastern Nile Delta

Article Title: Holocene landscape evolution and settlement development in the surroundings of the ancient Egyptian city of Bubastis (“Tell Basta”), southeastern Nile Delta

Article References: Garbe, P. (2026). Holocene landscape evolution and settlement development in the surroundings of the ancient Egyptian city of Bubastis (“Tell Basta”), southeastern Nile Delta. E&G Quaternary Science Journal, 75(2), 169-171. https://doi.org/10.5194/egqsj-75-169-2026

Image Credits: AI Generated

DOI: 10.5194/egqsj-75-169-2026

Keywords: Bubastis, Tell Basta, Nile Delta, geoarchaeology, Holocene, Geziras, phytoliths, sediment coring, electrical resistivity tomography, ancient Egypt, urban development, Nile flooding

The challenge the researchers set out to address is deceptively complex. Vegetation water stress expresses itself through coupled changes in canopy greenness, leaf area, surface temperature, and evaporative behavior, all of which can be observed from space. Satellite-derived indicators such as the Normalized Difference Vegetation Index (NDVI), the Enhanced Vegetation Index (EVI), Leaf Area Index (LAI), Fraction of Photosynthetically Active Radiation (Fpar), the Evaporative Stress Index (ESI), and Land Surface Temperature (LST) provide temporally consistent, spatially exhaustive information on crop condition. Yet translating these multiple, partly redundant signals into a reliable single estimate of vegetation stress has remained difficult, particularly in irrigated deltas where human water-management decisions strongly modulate how plants respond to climate forcing.

The study uses MODIS satellite products covering 2018 to 2025, with NDVI and EVI drawn from the MOD13Q1 product at 250-meter resolution, LAI and Fpar from MOD15A2H, LST from MOD11A2, and actual and potential evapotranspiration from MOD16A2, from which ESI was calculated as the ratio of actual to potential evapotranspiration. Quality-assurance layers were used to exclude pixels contaminated by clouds, shadows, aerosols, or low-confidence retrievals, and the data were aggregated into monthly governorate-level values. Observations from 2018 to 2023 served for model training, while 2024 and 2025 were reserved as an independent testing period, supplemented by 5-fold cross-validation to assess temporal robustness.

A central methodological innovation lies in the coupling of Best Subset Regression (BSR) with machine learning. Rather than feeding every available indicator into the models, BSR systematically evaluated all possible combinations of LAI, Fpar, ESI, NDVI, and LST, selecting the subset that best balanced explanatory power and parsimony using criteria such as adjusted R², the Akaike and Bayesian Information Criteria, and Mallows’ Cp. The Variance Inflation Factor was used to screen out multicollinearity. This physically grounded variable reduction was then paired with four machine-learning architectures: a multilayer perceptron (MLP) neural network, a Random Subspace (RS) ensemble, Regression by Discretization (RD), and Random Forest (RF).

The feature-selection analysis produced a strikingly clear hierarchy of controls. NDVI emerged as the overwhelmingly dominant predictor of EVI, with a correlation of 0.934 and a standardized regression coefficient of 0.935, supported by an extraordinarily large t-statistic of nearly 199 and a narrow, strictly positive confidence interval. LAI contributed a smaller but statistically significant refinement, capturing canopy structural attributes such as layering and leaf clumping that a two-band greenness index cannot fully represent. By contrast, Fpar, ESI, and LST showed no detectable linear contribution once NDVI and LAI were accounted for, suggesting that in this irrigated setting their influence is either mediated through greenness signals or expressed through nonlinear interactions, and that managed water supply buffers canopy condition from short-lived thermal or evaporative stress.

The consequences of feature optimization for model performance were substantial, though architecture-dependent. Before selection, Regression by Discretization led the independent testing period, while Random Forest lagged with a correlation of only 0.9571. After BSR, Random Forest transformed into the clear best performer, achieving a correlation of 0.9943, a mean absolute error of just 0.0063, and a root mean square error of 0.0161 during the 2024–2025 testing period, with relative absolute error falling to under 5 percent. Random Subspace also improved markedly, its testing correlation rising from 0.9414 to 0.9738 and its slope moving from 0.4768 to 0.9666, while the multilayer perceptron failed to benefit and deteriorated slightly across all metrics.

Cross-validation reinforced these findings. Across five folds, the optimized Random Forest achieved a correlation of 0.9793 with relative errors dropping to roughly 9.6 and 20.2 percent, demonstrating low bias and low variance across resampled data partitions. The authors attribute this behavior to the underlying data structure: because EVI is governed by a near one-dimensional NDVI–EVI relationship with a modest incremental contribution from LAI, tree-based ensembles can exploit repeated splits on these dominant predictors to form many shallow, low-bias trees, while bagging averages out noise. The neural network, with its larger parameterization relative to the effective information content, remained more sensitive to sample variation.

To confirm that the optimized Random Forest was not simply outperforming weak baselines, the team benchmarked it against univariate and multiple linear regression, XGBoost, and CatBoost. The linear baselines produced testing RMSE values of approximately 0.053, while the boosting algorithms reduced this to around 0.045 and 0.046. The feature-optimized Random Forest, however, retained a substantially lower RMSE of 0.0161, roughly a third of the boosting models’ error, confirming a genuine methodological advantage rather than an artifact of comparison. Time-series analysis further showed that after feature selection the model reproduced the timing and amplitude of seasonal greening and senescence cycles with markedly tighter agreement.

The practical implications extend to operational water management in semi-arid regions. Because NDVI and LAI are routinely available from satellite platforms at frequent intervals, the framework could be updated as new observations arrive, supporting governorate-scale vegetation-condition monitoring. Estimated EVI could then be converted into crop- and season-specific anomalies and combined with crop-water requirements and irrigation schedules to inform allocation decisions. The authors are careful to note limits: the framework estimates contemporaneous vegetation condition rather than forecasting future stress, and its findings are specific to Dakahliyah Governorate and the 2018–2025 period. Validation across longer records, contrasting irrigation regimes, and additional regions, along with the incorporation of microwave soil moisture and field-scale data, remains a priority before broader transferability or early-warning use can be claimed.

Nevertheless, the study offers a compelling demonstration that less can be more in agricultural machine learning. By systematically identifying which satellite indicators genuinely carry information about crop water stress, and by discarding the rest, the researchers built a model that is simultaneously more accurate, more interpretable, and cheaper to operate, requiring only a small set of widely available inputs. As water scarcity intensifies across semi-arid agricultural regions worldwide, such parsimonious, physically grounded AI frameworks may become essential instruments for safeguarding food production under increasingly uncertain climatic conditions.

Subject of Research: Estimation of agricultural vegetation water stress using feature-optimized remote sensing and ensemble machine learning in the Nile Delta

Article Title: Estimation of agricultural vegetation water stress using feature-optimized remote sensing and ensemble machine learning

Article References: Elbeltagi, A., Srivastava, A., & Alsumaiei, A. A. (2026). Estimation of agricultural vegetation water stress using feature-optimized remote sensing and ensemble machine learning. Smart Agricultural Technology, 15, Article 102536. https://doi.org/10.1016/j.atech.2026.102536

Image Credits: AI Generated

DOI: 10.1016/j.atech.2026.102536

Keywords: vegetation water stress, remote sensing, machine learning, random forest, Nile Delta, MODIS, EVI, NDVI, best subset regression, agricultural drought, irrigation management, Egypt