Recent decades have witnessed a danger to food security as well as to human health because of pollutants’ negative impact on crop quality. An accurate estimate of the heavy metal concentrations in Egypt’s north Nile Delta is required to lower the high concentration levels of heavy metal in the soil as a means to develop a remediation strategy that stabilizes heavy metals in contaminated soil. Using a geo-accumulation index (I-geo), contamination factor (CF), Improved Nemerow’s Pollution Index (Pn), and Potential Ecological Risk Index (PERI), supported by GIS; principal component analysis (PCA), and cluster analysis, six heavy metals (As, Co, Cu, Ni, V, and Zn) were analyzed from 15 soil profile layers (61 soil samples) to determine the extent of the soil contamination in the area studied. The findings demonstrate the widespread I-geo contamination of As, Co, Cu, Ni, V, and Zn in different layers. The ranges for the I-geo values were from −8.2 to 5.3; 4.11 to 1.8; 6.4 to 1.9; −9.7 to 2.8; −6.3 to 2.9; and from −12.5 to 2.4 for As, Co, Cu, Ni, V, and Zn, respectively. I-geo categorization therefore ranged from uncontaminated to strongly/extremely contaminated. The CF values varied from 0.01 to 60.6; 0.09 to 5.17; 0.02 to 10.51; 0 to 10.51; 0.02 to 7.12; and 0 to 7.68 for As, Co, Cu, Ni, V, and Zn, respectively. In decreasing sequence, the CFs are arranged as follows: CF (As), CF (Ni), CF (Zn), CF (V), CF (Cu), and CF (Co). Most of the research region (71.9%) consisted of a class of moderately to heavily polluted areas. Additionally, a large portion of the study region (49.17%) has a very high risk of contamination, as per the results of the PERI index. The use of a correlation matrix, cluster analysis, and principal component analysis(PCA) to evaluate the variability in the soil’s chemical content revealed the impact from anthropogenic activities on the heavy metal concentration levels in the study area’s soil. The current findings reflect the poor quality of management in the research region, which led to the increase in the concentration of heavy metals in the soil. Decision-makers could use the outcomes from the spatial distribution maps for contaminants and their levels as a basis for creating heavy metal mitigation strategies.

Land evaluation has an important role in agriculture. Developing countries such as Egypt face many challenges as far as food security is concerned due to the increasing rates of population growth and the limited agriculture resources. The present study used multivariate analysis (PCA and cluster analysis) to assess soil capability in drylands, Meanwhile the Almagra model of Micro LEIS was used to evaluate land suitability for cultivated crops in the investigated area under the current (CS) and optimal scenario (OS) of soil management with the aim of determining the most appropriate land use based on physiographic units. A total of 15 soil profiles were selected to characterize the physiographic units of the investigated area. The results reveal that the high capability cluster (C1) occupied 31.83% of the total study area, while the moderately high capability (C2), moderate capability (C3), and low capability (C4) clusters accounted for 37.88%, 28.27%, and 2.02%, respectively. The limitation factors in the studied area were the high contents of CaCO₃, the shallow soil depth, and the high salinity and high percentage of exchangeable sodium (% ESP) in certain areas. The application of OS enhanced the moderate suitability (S3) and unsuitable clusters (S5) to the suitable (S2) and marginally suitable (S4) categories, respectively, while the high suitability cluster (S1) had increased land area, which significantly affected the suitability of maize crop. The use of multivariate analysis for mapping and modeling soil suitability and capability can potentially help decision-makers to improve agricultural management practices and demonstrates the importance of appropriate management to achieving agricultural sustainability under intensive land use in drylands.

Developing countries all over the world face numerous difficulties with regard to food security. The purpose of this research is to develop a new approach for evaluating wheat’s suitability for cultivation. To this end, geographical information systems (GIS) and fuzzy inference systems (FIS) are used as the most appropriate artificial intelligence (AI) tools. Outcomes of investigations carried out in the western Nile Delta, Egypt. The fuzzy inference system used was Mamdani type. The membership functions used in this work are sigmoidal, Gaussian, and zmf membership. The inputs in this research are chemical, physical, and fertility soil indices. To predict the final soil suitability using FIS, it is required to implement 81 IF-THEN rules that were written by some experts. The obtained results show the effectiveness of FIS in predicting the wheat crop’s suitability compared to conventional methods. The research region is split into four classes: around 241.3 km² is highly suitable for wheat growth, and 224 km² is defined as having moderate suitability. The third soil suitability class (low), which comprises 252.73 km², is larger than the unsuitable class, which comprises 40 km². The method given here can be easily applied again in an arid region. Decision-makers may benefit from the research’s quantitative findings.

Nowadays, machine learning (ML) is a useful technology due to its high accuracy in constructing non-linear models and algorithms that can adapt to the complexity and diversity of data. Thus, the current work aimed to predict the soil quality index (SQI) from extensive soil data, achieving high accuracy with the artificial neural networks (ANN) model. However, the efficiency of ANN depends on the accuracy of the data that is prepared for training. For this purpose, MATLAB programming language was used to enable the calculation, classification, and compilation of the results into databases within a few minutes. The proposed MATLAB program was highly efficient, accurate, and quick in calculating soil big data for training the machine compared with traditional methods. The database contains 306 vector sets, 80% of them are used for training and the remaining 20% are reserved for testing. The optimal model obtained comprises one hidden layer with 250 neurons and one output layer with a sigmoid function. The ANN achieved a high coefficient of determination (R²) values for SQI estimation, with around 0.97 and 0.98 for training and testing, respectively. The results indicate that 36.93% of the total soil samples belonged to the very high quality class (C1). In contrast, the high quality (C2), moderate quality (C3), low quality (C4), and very low quality (C5) classes accounted for 10.46%, 31.37%, 20.92%, and 0.33% of the samples, respectively. The high contents of CaCO₃, pH, sodium saturation, salinity, and clay content were identified as limiting factors in certain areas. The results of this study indicated high accuracy of soil quality assessment using physical, chemical, and fertility soil features in regression analysis with ANN. This method, which is suitable for arid zones, enhances agricultural productivity and decision-making by identifying critical soil quality categories and constraints.

Evaluating soil quality is crucial for ensuring the sustainable use of agricultural lands. This review examines the definition, evaluation methods, indicator selection, and relevant case studies. The concept of soil quality supplements soil science research by deepening our understanding of soils and aiding in the allocation of resources as agriculture intensifies to meet rising global demand. Soil quality provides a framework for educating stakeholders about the essential functions of soils and offers a tool for assessing and comparing different management techniques. Regular evaluation of soil quality is vital for maintaining high crop yields and addressing the gap between production and consumption. Nowadays, many researchers have explored machine learning (ML) and deep learning (DL) techniques and various algorithms to model and predict soil quality with satisfactory results. These chosen indicators can be influenced by chemical, biological, or physical features. This paper compares ML and DL with traditional methods, examining their features, limitations, different categories of machine learning, and their applications in soil quality assessment. Finally, we show that predicting soil quality has the potential to be extremely accurate and efficient with ML and DL. This distinguishes the application of DL and ML from other approaches since they can anticipate the soil quality index without the need for more intricate computations. Our suggestion for future studies is to evaluate soil quality over broader regions and predict it by using more accurate, modern, and faster methods, using a variety of activation functions and algorithms.