Piezometric depth modeling of groundwater using monthly variables of precipitation and water consumption (case study: Sarab Plain aquifer)

Abstract

AbstractThe growth of the world population and the problem of food supply have led to the development of agricultural land, particularly in the Third World and in Iran, and thus to a sharp increase in water consumption regardless of the existing water resources. On the other hand, the ever-increasing growth of industries and factories, regardless of the impact on the environment, together with the increase in water consumption, has disturbed the balance of the environment and caused climate change with rising temperatures and increasing pollution. Unfortunately, the management of water resources and the environment is incompatible with the development of agricultural land and the development of industries, and therefore in most countries of the world a situation has arisen in which surface and groundwater resources are at risk. The two main variables, precipitation and water consumption, control groundwater levels. The area studied in this research is the Sarab Plain aquifer located in East Azerbaijan province, Iran. In the Sarab Plain and other plains of Iran, indiscriminate harvesting has led to a significant decline in the groundwater level (in other words, piezometric depth) and subsidence of the plain. The area under cultivation of various agricultural crops such as beans, cucumbers and alfalfa and gardens is about 38,176 ha, irrigated by 739 licensed wells. Agricultural uses on the one hand and industrial and animal uses on the other led to a progressive lowering of the piezometric level of the plain. The average water consumption from the table is currently 53 million cubic meters per year, while the amount of renewable water is 35.81 million cubic meters per year. The data used in the study are monthly precipitation from 19 rain gauge stations, monthly piezometric codes from 78 piezometers converted to piezometric depth, and monthly water consumption from 1886 consumable wells between 2007 and 2022. Individual regression relationships were created between the piezometric depth variable and consumption and precipitation variables. In the first step, a general hybrid exponential relationship between piezometric depth, consumption and precipitation was found. The correlation coefficient value between the calculated and observed piezometric depth was 0.69. Furthermore, the root mean square error and Kling-Gupta were 2 m and 0.57, respectively. In order to apply the hybrid exponential relationship to predict piezometric depth in the coming years, it was necessary to predict precipitation and consumption. To predict monthly precipitation based on its periodicity, the Thomas and Fiering (T&F) consumption forecasting method was used. 20% of the data was compared with calculated data. The result showed, R = 0.815 and RMSE = 0.07 mm between calculated and observed data. Additionally, to predict consumption in the coming years, a suitable regression relationship between consumption and time was constructed, showing a correlation of 0.97 and a root mean square error of 0.0008 mcm with observations. In the second step, precipitation and consumption were predicted for the next 3 years (2023–2025) and piezometric depth were determined for this period by applying them in the hybrid model. The forecast for the next three years shows that the upward trend of the piezometric level will continue. The application of the regression method resulted in a final equation, which is particularly important in view of the stabilization of the piezometric level of the reservoir. This method has no particular limitations and is an appropriate method when accurate consumption water and precipitation statistics are available. The only limitation that can be considered with this method is the movement around the average values and does not take into account the positional fluctuations. This work is new because it calculates groundwater simultaneously using two parameters: precipitation and water consumption. Other similar studies did not use groundwater consumption data.