Factors Affecting Domestic Water Consumption on the Spanish Mediterranean Coastline

TLDR: In the last decades on the Spanish Mediterranean coastline there has been a great development of low-density urban areas, as well as a change in the sociodemographic structures, especially in t... as mentioned in this paper.

Abstract: During the last decades on the Spanish Mediterranean coastline there has been a great development of low-density urban areas, as well as a change in the sociodemographic structures, especially in t...

Figures

Table 4. Moran's I applied to model variables and regression residuals

Table 3. Results of the OLS regression model

Table 2. Descriptive statistics on dependent and independent variables

Table 6. Results of the geographically weighted regression model

Table 1. Dependent and independent variables description

Figure 2. Mapped GWR local R² results in the Spanish Mediterranean coastline municipalities

Full text

For Peer Review Only
Factors
affecting domestic water consumption in the
Spanish Mediterranean coastline
Journal:
The Professional Geographer
Manuscript ID
PG-2017-0058.R3
Manuscript Type:
Revised Manuscript
Key Words:
Domestic water consumption, determinants, tourism, Spain, Mediterranean
Abstract:
During the last decades in the Spanish Mediterranean coastline there has
been a great development of low density urban area, as well as a change
of the sociodemographic structures, especially in the municipalities that
have developed a residential tourism model. Likewise, urban and tourist
development have stressed the balance between the availability of water
resources and urban water demands, generating situations of scarcity
which may be aggravated by climate change. This study identifies the
determinants of water consumption in the Spanish Mediterranean coastline,
focusing on the variables related to urban land uses, socioeconomic and
sociodemographic variables at the municipal level using an ordinary least
square (OLS) and a geographically weighted regression (GWR) model. The
GWR model results substantially improved the results of the OLS model,
explaining 88.27 percent of the variance in domestic water consumption
and solving the spatial autocorrelation problem of some independent
variables. The most influential variables include the percentage of second
homes or the percentage of residential properties with swimming pools at
municipal level. These characteristics must be considered to develop
demand management policies and an updated water planning to ensure
urban supply in a future with less available water resources.
This is a previous version of the article published in The Professional Geographer. 2018, 70(3): 513-525. doi:10.1080/00330124.2017.1416302

For Peer Review Only
1
Factors affecting domestic water consumption in the Spanish Mediterranean coastline
Rubén Alejandro Villar-Navascués
University of Alicante
Alfredo Pérez-Morales
University of Murcia
Page 1 of 54

For Peer Review Only
2
Abstract: During the last decades in the Spanish Mediterranean coastline there has been a
great development of low density urban area, as well as a change of the sociodemographic
structures, especially in the municipalities that have developed a residential tourism model.
Likewise, urban and tourist development have stressed the balance between the availability of
water resources and urban water demands, generating situations of scarcity which may be
aggravated by climate change. This study identifies the determinants of water consumption in
the Spanish Mediterranean coastline, focusing on the variables related to urban land uses,
socioeconomic and sociodemographic variables at the municipal level using an ordinary least
square (OLS) and a geographically weighted regression (GWR) model. The GWR model
results substantially improved the results of the OLS model, explaining 88.27 percent of the
variance in domestic water consumption and solving the spatial autocorrelation problem of
some independent variables. The most influential variables include the percentage of second
homes or the percentage of residential properties with swimming pools at municipal level.
These characteristics must be considered to develop demand management policies and an
updated water planning to ensure urban supply in a future with less available water resources.
Key Words: Domestic water consumption, determinants, tourism, Spain, Mediterranean.
Page 2 of 54

For Peer Review Only
3
In Spain, the explosion of the real estate bubble and the economic crisis of 2008 have had
serious consequences in the urban and demographic dynamics, ending the intense property
development and population growth. This fact accentuated the regressive trend in urban water
consumption, which began some years earlier in some cities along the Spanish Mediterranean
coastline (Gil et al. 2015). Various reasons have been attributed to this reduction, such as
technical improvements in urban water management (Morote, Hernández and Rico 2016); the
reduction of disposable income by households (March and Saurí 2017); the increase in water
rates and prices (Morote, Hernández and Rico 2016); the change in the demographic structure
(March, Perarnau and Saurí 2012), or the impact that the restrictions of certain domestic uses
in drought situations have had on the modification of the consumption habits (March,
Domènech and Saurí 2013). However, an analysis of the explaining factors of domestic water
demand for the Spanish Mediterranean coastline has not yet been carried out, which may
contribute to explain this trend.
Together with the studies of the Anglo-Saxon tradition on factors affecting domestic water
demand (March and Saurí 2009), numerous studies have been carried out in Europe in the last
decade (Mazzanti and Montini 2006; Romano, Salvati and Guerrini 2014) and more
specifically in the Spanish Mediterranean coastline (Domene and Saurí 2006; March and
Saurí 2010; March, Perarnau and Saurí 2012; Garcia, Llausàs and Ribas 2013). The research
carried out has confirmed that one of the main determinants of domestic water consumption
is the predominant urban form, due to the influence that external uses have on final
consumption (Wentz and Gober 2007; March and Saurí 2010; House-Peters, Pratt and Chang
2010; Morote, Hernández and Rico 2016). Also, other Spanish studies have focused on the
influence that the tourism sector exerts in the increase of the domestic demand of water, due
to the impact of detached houses for residential tourism use on the seasonal water
consumption (Rico 2007; Rico, Olcina and Saurí 2009; Hof and Schmitt 2011; Hof and
Page 3 of 54

For Peer Review Only
4
Blázquez 2013, 2015; Hof and Wolf, 2014). However, according to the literature consulted,
the influence of tourism on domestic water demand has not been analyzed at international
level, since research has focused mainly on water consumption by hotels or leisure activities
(Gopalakrishnan and Cox 2003; Gössling et al. 2012; Gabarda-Mallorquí, Garcia and Ribas
2017).
Our hypothesis is that domestic water demand on the Spanish Mediterranean coastline is not
only determined by urban factors, but also by the residential tourism specialization of the
municipality and the sociodemographic and socioeconomic characteristics of the population.
The aim is to ascertain the influence of these variables on domestic water consumption,
measured in liters / inhabitant / day, through an ordinary least square (OLS) regression model
and a geographically weighted regression (GWR). Considering the future reduction of
available water resources because of climate change (IPCC, 2013) measures are needed to
address water scarcity (Jaeger et al. forthcoming), but the budgetary constraints of public
administrations make it difficult to maintain the supply-oriented water policies, based on
engineering solutions and the construction of large water infrastructures (Arahuetes, Villar
and Hernández 2016). These circumstances require the adoption of demand-oriented
measures, for which a better knowledge of the factors that affect domestic water demand is
necessary. It should be borne in mind that demand management measures are usually focused
on the increase of rates and the price of water, which can lead to the emergence of water
poverty situations even in countries of the so-called developed world (March and Saurí
2017). Thus, the results of this work are oriented to improve other demand management
measures such as the design of awareness campaigns, hydrological planning and other
measures of demand management beyond economic ones.
The article is organized as follows. First, we introduce a brief literature review on the main
factors influencing domestic water consumption, with a special emphasis on urban-tourist and
Page 4 of 54

Frequently asked questions

Q1. What have the authors contributed in "Factors affecting domestic water consumption in the spanish mediterranean coastline" ?

This study identifies the determinants of water consumption in the Spanish Mediterranean coastline, focusing on the variables related to urban land uses, socioeconomic and sociodemographic variables at the municipal level using an ordinary least square ( OLS ) and a geographically weighted regression ( GWR ) model. 

Q2. What are the future works mentioned in the paper "Factors affecting domestic water consumption in the spanish mediterranean coastline" ?

The use of aggregate city-scale data has made it difficult to study climatic variables, which, according to the international literature, are the main variables to explain the differences in domestic consumption between the base use ( winter ) and the seasonal use ( summer ) ( House-Peters and For Peer Review O nly 18 Chang 2011a ; Chang et al. 2017 ). Nonetheless, to carry out this analysis in tourist environments, it will be necessary to calculate the temporary population since this variable affects water consumption more than the climate. 

Q3. What is the main purpose of the regression analysis?

In the regression analysis, the logarithmic form of domestic water consumption (LNDWC12) is used to give robustness to the model, to comply with normalcy tests (Kolmogorov–Smirnov and Shapiro–Wilk) and to facilitate the interpretation of the results (March, Perarnau and Saurí 2012). 

Q4. Why did it be decided to carry out the analysis for 2012?

Due to the lack of homogeneous data series, it has been decided to carry out the analysis for the year 2012, as are available more records. 

Q5. How many municipalities are included in the sample?

It should be noted that although the sample represents 46 percent of the total coastal municipalities, it incorporates 56.2 percent of municipalities with more than 5.000 inhabitants, and concentrates 71 percent of the population registered in 2012 (INE 2013a), so this analysis is being carried out for the most populated municipalities. 

Q6. What is the main reason why the GWR results are not in line with other studies?

the results of the GWR indicate that it is necessary to develop further analysis on the local scale since it has been observed that the explanatory power of income level varies spatially. 

Q7. What is the main limitation of the study?

Another limitation is the use of aggregate city-scale data on domestic water consumption, because the authors have assumed "lack of variation in spatial patterns and processes" (House-Peters and Chang 2011a:4). 

Q8. What is the main reason why the results of the OLS model are improved?

the results of the OLS model are improved since it is possible to differentiate the local spatial variations of the parameters estimated by means of the implementation of a kernel function, that allows to make estimations adjusted to each observation giving greater influence on the closer observations (Brunsdon, Fotheringham and Charlton, 1996). 

Q9. What is the effect of the temporary population on the income variable?

In these cases, the temporary population influences the increase of the dependent variable, especially in municipalities with a high percent of second homes, and the underestimation of the income level, since it is calculated from the permanent population. 

Q10. What are the limitations of the study?

the authors recognize the limitations of the study, mainly due to the lack of rigorous information on the temporary population and the length of their stay.