چكيده لاتين :
Introduction
Groundwater resources are the most important and the cheapest water resources
the proper recognition of which and their exploitation principles can lead to
sustainable development and economic activities, particularly in arid/semi-arid areas.
The groundwater level reduction and human activities are reducing the groundwater
resources level. Fluctuations study are the most important management strategies
can be useful in this case. Parsabad of Moghan plain due to its form and geomorphic
structure is a flat plain. By considering the slope, elevation, quaternary sediments, old
and new fluvial terraces and alluvial fans, groundwater levels have fluctuated a lot.
At the same time, this fluctuation has taken a positive trend in recent years.
Therefore, groundwater level in the plain has reached near the surface of the Earth
caused many problems for people. Consequently, swamp lands and salinization of
agricultural land, destruction of urban and rural buildings and sinking the roads and
buildings may be caused due to the problem. So, to overcome this status, it seems
necessary prediction of the groundwater level fluctuations accurately.
Material and Methods
Obtaining a suitable model to predict the behavior of water resource taking into
account the factors affecting these phenomena is one of the most important
approaches in water resource management. These important factors include slope,
elevation, porosity, permeability, transmissibility, precipitation and
evapotranspiration. Therefore, in this study, observational wells data, meteorological
stations data, geological data, topographic maps and satellite images of ETM+ have
been used for variables mapping. Influenced factors maps is provided in GIS using
digital elevation model and Kriging. The effects of geographical factors on fluctuating
groundwater level have been explored by Pearson correlation. After determining the
affected factors on fluctuations of groundwater level, its trend had been anticipated
by taking these factors and multivariate linear regression model. In this context,
through analysis of variance {ANOVA), sum of squares, degrees of freedom, mean
square, Fisher's exact test and a significance level of regression have been examined.
Discussion
The results showed slope, elevation, transmissibility and porosity and in contrary, the
permeability, precipitation and evaporation are in high and low significance level of
correlation with fluctuating groundwater levels, respectively. Also, no significant
correlation is between groundwater depths and groundwater level, therefore, have
no effect on the groundwater level fluctuations. The results implied the consistencywith reality about the direct/indirect influence of considered environmental factors in
the groundwater level fluctuation. So, the groundwater level have been decreased by
increasing slope, height, transmissibility and evapotranspiration. But, the porosity,
permeability and precipitation have a direct relationship with groundwater level.
Standard deviation parameters, standardized coefficients, t test, and significance
level have been investigated in regression model. The significance level (Sig) is 0.000
or near to zero in all parameters. Thus, the model implemented in this study have
been significant, so, prediction map of groundwater level fluctuations based on the
model have been produced by spatial analysis and overlapping raster.
Conclusion
Due to the region characteristics, slope and elevation are the most important factors
affected the groundwater level fluctuation. Since the slope of study area is close to
zero in the northeast regions and the Beach of Aras River, so the motion of
groundwater in these areas have been very slow. On the other hand, the formation of
this area is clay and silt causes slowing down the movement due to their heavy soils
characteristics. Beside the geographical features affected in groundwater fluctuating,
extensive networks of water supply with canal to irrigate the agricultural areas are
more important factor, too. For evaluating the model, relationships between
estimated and observed groundwater level have been investigated by correlation,
RMSE and R2
• In this study, these criteria implied R = 0.988, RMSE = 0.07 and R2 =
0.9712, so, the results reflected high efficiency of the process model of groundwater
level fluctuations. Due to the lack of these models in the country, this study can be
used as a prelude to develop prediction models of groundwater level.
