Use of Geomatic Techniques to Determine the Influence of Climate Change on the Evolution of the Doñana Salt Marshes’ Flooded Area between 2009 and 2020

Abstract

Located in the south of the Iberian Peninsula, the Doñana salt marshes occupy around half of Doñana National Park and are currently considered among the most important wetlands worldwide due to the importance of their ecosystem. In this research work, using a novel patented procedure, the effects of climate change on the study area between 2009 and 2020 were evaluated. For this reason, DEMs were downloaded from the 30-meter Shuttle Radar Topography Mission (SRTM). Furthermore, to check the depth of the flooded area, 792 satellite images (L5 TM, L7 ETM+, and L8 OLI) with a resolution of 30 m were analyzed. The results show how the combined use of geomatic techniques, such as radar, optical, and geographic information system (GIS) data, along with regression models and iterative processes, plays a key role in the prediction and analysis of the flooded area volume in the Doñana salt marshes. Another significant contribution of this work is the development of a new remote sensing index. In conclusion, given that the study area depends on its aquifers’ status, it would be advisable to implement policies aimed at eradicating illegal aquifer extraction, as well as recovery plans to avoid the complete clogging of this salt marsh.

1. Introduction

Paleogeographic studies have shown that climate change is inherent to the Earth’s evolution and a constant throughout human history [1]. In a broad sense, human activity began transforming the Earth’s surface with the spread of agricultural production at the beginning of the Neolithic period [2]. This transformation became particularly significant following the Industrial Revolution, influencing aspects such as biodiversity loss, the increase in desertification processes, and the current characterization of the planetary climate. These changes deviate from the natural evolution of the Earth in both genesis and intensity. Instead, they are the direct consequences of disruptions to the environment caused by human activity, leading to what is referred to as (hu)man-induced climate change [3].

Coastal areas are crucial in climate change studies as they provide information on potential risks and impacts, as well as the measures to be considered for mitigating or eliminating them [4]. For this reason, it is necessary to consider these areas globally, that is, as a varied environment including transition zones such as coves, gulfs, and estuaries, since their influence on both hydrodynamics and morphodynamics is of great ecological interest [5].

In relation to the above, the Doñana marshes, located in southwestern Spain, are part of a larger functional entity: the extensive coastal ecosystem of the Doñana. This area represents one of the most significant European fluvio-coastal complexes, almost entirely situated within Doñana National Park. According to Paredes et al., this complex comprises several eco-districts (lower-ranking systems) identified through lithological and morphodynamic analyses [6], specifically categorized as follows: aeolian (comprising active, semi-stable, and vegetated dunes); coastal (consisting of the Las Marismillas barrier island, along with both ancient and current beaches); Condado sandfields; sandy continental terrains; and, finally, silt and clay marshes resulting from the infilling of the ancient Guadalquivir estuary.

Water resources, including surface and groundwater, are crucial factors in these protected areas [7] as they form the strategic foundation for maintaining ecosystems in general and particularly in marsh environments, lagoon systems, and river streams. However, in recent decades, due to the progressive temperature increase and precipitation decrease along the southern Iberian Peninsula, the floodable area of the Doñana marshes has significantly decreased [8,9], leading to a lower volume of stored water, resulting in a decreased marsh sediment renewal rate, which has impacted various plant and animal communities [10].

According to León-Palmero, limnology plays a crucial role in decision making regarding the management of impounded water [11]. Therefore, conducting marsh studies from a multifactorial perspective, considering their ecology, physico-chemical aspects, and biological characteristics, is of great interest.

While various field studies have been conducted to examine ecosystem processes and their implications for mitigating future global land degradation [12] or on how to evaluate a population’s high flooding vulnerability due to climate change in wetland ecosystems [13], it remains difficult to generalize findings from patch-scale research to other regions. In the field of geomatics, while satellite platforms can help us address this issue, light detection and ranging (LIDAR) data have a clear benefit, as they precisely capture marshes’ physical characteristics, which traditional contour mapping often fails to achieve [14,15]. Consequently, LIDAR data provide the crucial spatial information needed for a detailed marsh-scale analysis of the relationships between climatic variables and soil erosion. LIDAR technology scans its field of view with one or more laser beams to measure the distance to objects [16]. Sometimes, a single light pulse can produce multiple reflections, such as when passing through tree canopies. LIDAR systems can collect data from a canopy’s top layer to the ground [17], creating a more accurate digital elevation model (DEM) by removing vegetation.

From a global point of view, certain studies [18,19,20] are noteworthy as they establish the basis for understanding salt marsh erosion due to climate change. Recently, several researchers have focused on the dynamics of salt marshes in the Iberian Peninsula [21,22,23,24]. In the Doñana marshlands (Huelva, Seville, and Cádiz Provinces, Spain), sediment residence times are prolonged, with a significant intermediate storage of eroded material, a common trait among major marsh systems worldwide, such as those in Chesapeake Bay (USA) [25]. Additionally, increased sediment storage can lead to major changes in marshland systems’ physical structure and ecological health, crucial for understanding the flora and fauna in these areas. Another study used geomatic techniques to assess the dynamics and evolution of an Asturian coastal sandy system in northern Spain [26].

Another effect of long sediment residence times in river systems is that significant historical changes can impact a system’s behavior for decades [14,25,27]. Understanding and assessment of a system’s response to historical global changes are essential, as the temporal scale of analysis is vital for forecasting how the net sediment response will relate to flood distribution over time.

Hitherto, most marshland studies have concentrated on understanding soil loss and erosion through mapping vegetation cover and general land use changes (LUCs). Some researchers have investigated various methods for determining variable states using soil reflectance or vegetation cover characteristics [12]. Additionally, LUCs have been used to analyze watershed dynamics [28] and assess fluvial sediment transport degradation [29]. The number of published articles on the innovative use of LIDAR images in this research area is relatively low, despite its significant potential, possibly due to the fact that it is more difficult to obtain at various scales compared to satellite images, consequently favoring remote sensing for marshland studies.

Moreover, utilizing LIDAR techniques alongside iterative scientific methods and fieldwork to quantify erosion processes can offer new insights for predicting the total volume of eroded soil in wetlands and other regions globally. Notwithstanding, high-resolution LIDAR data are essential for areas with complex networks of basins or aquifers. A promising approach is the use of emerging image processing techniques [30] this being increasingly recognized by researchers worldwide.

Nevertheless, it is worth noting that the acquisition of DEMs at a global scale using radar has a much longer historical trajectory, as evidenced by the use of DEMs generated from the NASA Shuttle Radar Topography Mission (SRTM), whose first version dates back to the year 2000 [31].

In this regard, this research aims to forecast the average soil deposition in the full area of the Doñana salt marshes, using a novel approach involving DEM data and geographic information systems (GISs). To accomplish this primary goal, predicting volumetric change processes is an essential secondary objective of this research, alongside the development of a new remote sensing index and algorithm.

Although DEMs obtained from either LIDAR or drone data (with scarce temporal coverage or limited spatial and temporal coverage, respectively) are currently used, this study was conducted using DEMs from the SRTM, with a resolution of 1 arc-sec (approximately 30 m) and a major temporal extent [31]. Additionally, it is important to highlight that this study’s novelty lies in the use of a patented procedure that allows us, using climatic data from the study period, to infer changes in the Doñana marshlands through big data analysis. It is also crucial to consider various works conducted by other researchers [21,24,32,33], focusing on coastal gulf areas in Texas (USA) and Cádiz (Spain).

2. Study Area

2.1. Study Area and General Characteristics

Located in the southwest of the Iberian Peninsula, the Doñana marshes are within the Doñana Natural Space, which spans the provinces of Seville, Huelva, and Cádiz in Spain and includes both Doñana Natural Park and Doñana National Park (Figure 1).

According to Green et al., until the end of the 19th century, the Doñana marshes covered 140,000 hectares [35]. Since then, however, due to various and intense human activities (such as agriculture, aquaculture, and tourism, among others related to climate change), they have been reduced to about 33,000 hectares of semi-virgin marshes. Given their relevance as one of the world’s most significant wetlands due to their rich wildlife, hosting over six million wintering birds annually, the Doñana salt marshes were designated a national park in 1960 (Doñana National Park), a UNESCO World Biosphere Reserve in 1980, and a Wetland of International Importance (Ramsar Convention) [35].

2.2. The Development and Evolution of the Doñana Salt Marshes

The Doñana marshes are the result of the gradual evolution of the ancient estuary of the Guadalquivir River, now isolated from the sea by a complex and significant system of coastal barriers that developed throughout the Holocene [36]. Their geomorphology is influenced by the evolution of the river network, composed of a series of fluvial levees delineating depressed areas and acting as natural barriers, regulating the water levels supplied by the river network and preventing the entry of tidal flows.

In their final evolutionary stages, their dynamics were influenced primarily by rainfall. From a climatic perspective, the Doñana has a mild Mediterranean climate, its most characteristic feature being its high seasonality, with a rainy period from October to April and a dry period from May to September. The average annual precipitation is around 537 mm, although there have been periods with less than 300 mm (years 1980 and 1994) and others exceeding 900 mm (years 1987 and 1995) [37]. From a fluvial perspective, the Guadalquivir and Guadiamar rivers are the main watercourses influencing the hydrodynamics of this area, with average annual flows of 164 m³/s [38] and 6.6 m³/s [35], respectively. These flows are affected by strong climatic seasonality, with minimum values in the summer months and maximum values in the winter, as well as interannual irregularity. During rainy winters, the flows can exceed 1000 m³/s, causing major floods. Currently, the tidal influence within this marsh area is nonexistent [35], although it was very significant during the Holocene.

2.3. The Climatic Importance of the Doñana Salt Marshes

Concerning the study area’s climatic evolution, precipitation and temperature variables from 2006 to 2100 were considered. As a starting point, the average values of these variables were considered under the SSP3-7.0 Scenario and the MCG CMCC-ESM2 mathematical model from the Local Climate Change Scenarios of the Andalusian Regional Government, based on the 6th report of the Intergovernmental Panel on Climate Change (IPCC) [39].

It is crucial to underscore that the study area depends not only on the climate to maintain its ecological–environmental value, but also on the contribution of subsurface runoff that reaches the marsh through the aquifers. For this reason, in this work, taking into account both the geology and hydrogeology of the Doñana salt marshes is key.

2.4. The Geological and Hydrogeological Importance of the Doñana Salt Marshes

The Doñana marshes are located within the aquifer subunit known as Almonte–Marismas, covering approximately 2900 km², composed of sediments from the Upper Miocene to the Lower Pliocene [40]. The entire formation lies almost horizontally, unaffected by significant tectonic movements. Specifically, the aquifer’s formation occurred when the sediments that constitute it were deposited over a layer of blue marls and sandy silts in fluvio-marine processes. This created a 300 m deep layer of unconsolidated fine materials, up to 2000 m in total thickness. These materials act as an impermeable boundary and outcrop to the north of the aquifer. Their primary origin is the southern part of the Iberian Massif, composed of metamorphic rocks, resulting in the presence of quartz, clays, and feldspars in the marsh area.

It is important to note that, until the Pliocene [41], the study area was submerged, and, during this period, the sea began to retreat, facilitating the deposition of sands, which later emerged and underwent alteration, and upon which gravels were subsequently deposited.

Additionally, the presence of water in the Doñana marshes is mainly due to contributions from the aquifer, the recharge of which is favored by the sandy areas surrounding the marshes, resulting in minimal surface runoff. This aquifer discharges continuously into streams in a diffuse manner [40].

However, over time, there will be a progressive decrease in the water volume responsible for recharging the aquifer, crucial for flooding the Doñana marshes, which will have severe environmental consequences, a fact which must be taken into account when aiming to minimize human impact, i.e., “(hu)man-induced climate change” [3].

3. Methodology

The methodology employed in this investigation consisted of multiple sequential procedures (Figure 2).

In this work, digital elevation models (DEMs) for different time periods (namely, 2009, 2013, and 2020) were first acquired for the studied area. Subsequently, control sites were selected, deriving their heights from the DEMs. Later, a genetically changed algorithmic model was trained using only 1% (17) of the total 1700 randomly picked elevation data points previously collected. The smallest number of randomly selected points was chosen, yielding the same result with 99% accuracy [30], taking into account variations in land usage. Then, the model, designed to provide a mathematical equation for forecasting the water mass volume, was assigned the responsibility of calculating the expected results at many randomly chosen locations evenly distributed around the research region. At this point, the result was a calculated approximation of the volume. After considering the climate variables, the rainfall amounts, and the randomly selected sites, the model was tasked with conducting both past and future sediment deposition simulations. Subsequently, the model utilized 1700 randomly obtained data points to forecast the alterations in both the surface and volumetric properties across the study period. Ultimately, the model predicted the mean clogging process in the marshland through the use of a ground shape factor (slope angle-dependent), based on both the shadow cast per pixel obtained in the period with no flooding and the elevation of the control points iterated across all the images in the evaluated time series.

To check the obtained results’ coherence, a comparison was carried out with what had been inferred for the years 2008/2009, 2009/2010, 2010/2011, 2011/2012, 2012/2013, 2013/2014, 2014/2015, 2015/2016, 2016/2017, 2017/2018, 2018/2019 and 2019/2020 through the use of 792 satellite images with a 30 m resolution. Among these, 330 came from Landsat 5 TM (Thematic Mapper), 342 from Landsat 7 ETM+ (Enhance Thematic Mapper Plus), and 120 from Landsat 8 OLI (Operational Land Imager). All the images had less than 20% cloud cover. It is relevant to note that the flood cycle begins in October/November of the previous year and usually ends in February/March of the following year.

Data Source Description

To understand DEMs, it is essential to emphasize their importance in accurately predicting climate change-linked soil deposition. Recognized as vital global geospatial data, DEMs are endorsed by prominent organizations such as the Spanish National Cartographic System, the INSPIRE Directive, and the United Nations [42]. In Spain, the national mapping agency, the Instituto Geográfico Nacional (IGN), has utilized LIDAR data from the PNOA photogrammetric flight to create the digital elevation model [34]. At a global scale, since 2000, NASA has conducted various SRTMs to determine global DEMs, with a grid spacing of 1 and 3 arc-seconds. In the current research, these DEMs were downloaded for the time range 2009–2020 from NASA’s Earth Explorer (https://earthexplorer.usgs.gov/, accessed on 20 June 2024).

In this work, we strategically placed 17 control points throughout the study area (Figure 3).

DEM data from NASA’s SRTM were augmented with the chosen control points and used to calculate the total water volume in the study area via iterative processes [43]. Likewise, an iterative data processing method was used to ascertain the average water depth, which was essential for gathering data on soil erosion in the Doñana salt marshes.

Next, a thorough literature review was conducted to determine the potential availability of a sufficiently long precipitation record for deriving relevant information for this research [44]. In this context, based on [45], precipitation (daily rainfall data) and temperature (daily data) records (from 2006 to 2020) were obtained, enabling the extraction of variables (Figure 4) pertinent to the Doñana salt marshes during the research period from 1 October 2008 (the start of the flooding cycle of 2008/2009) to 31 May 2020 (the end of the flooding cycle of 2019/2020).

Furthermore, utilizing data mining to analyze soil deposition between 2009 and 2020 involved extracting significant information from extensive datasets to reveal patterns and trends, which was then organized in a comprehensible way for further use by modified genetic algorithms. These modified genetic algorithms [46] work as a parallel process, operating as follows: Initially, the extensive dataset to be categorized is stored in a central storage unit. Upon execution, individual segments are processed through separate mapping tasks, each one beginning by accessing the training dataset and training the classifier. Then, the trained classification model is used to classify the large dataset. The iterative training process for each mapping task is designed to minimize the computational impact, as the training dataset is small compared to the extensive one, which accounts for most of the computing time.

4. Results

4.1. DEM Usage and Land Use Changes

The initial step involves obtaining a DEM. In this study’s instance, three DEMs were obtained from SRTM data (from 2009, 2013, and 2020) to coherently explain the erosion processes that occurred.

Regarding land use changes (LUCs) in the Doñana marshes (Figure 5), the primary land use characteristic in the study area at the beginning of 2016 was the predominance of natural and forested spaces, occupying 68% of the territory [47]. As the main changes in anthropic actions arose after 2016, this is considered a reference year. It is important to highlight that this area has traditionally lacked significant economic value because the locals focused on its utility value as a resource to increase income [48].

In addition, as a result of state initiatives [47], there has been an economic model shift through the promotion of large forested areas to meet the local population’s needs. Therefore, the predominant land uses observed in this area are wooded formations, with or without scattered natural shrubland, comprising 30% of the total. Considering the previously mentioned limited economic value, the extensive open spaces with sparse vegetation or occupied by shrub or herbaceous associations (approximately a quarter of the total surface area) are worth noting. Completing the spatial distribution of land uses are the rainfed crops associated with the Mediterranean trilogy of olive, vine, and cereal, the wetland mass formed by the Doñana salt marshes, and the prior flooding of the Rocina area adjacent to them. Finally, it is important to point out that, although irrigation does not exceed 2% of coverage, and given that the filling of the Doñana marshes is mainly due to rain and aquifer contributions, it is essential to consider it a key factor in the global warming of the surrounding area [47].

4.2. Prediction of Flooded Area and Volume

After examining all the available data, it was observed that there had been a progressive decline in surface water in the research area due to an ongoing dry spell. This was similar to the GIS study in [50], which correlated the DEM of a natural area with the region’s meteorological features. This analysis aimed to predict the average flooded surface volume in cubic hectometers, following the implementation of multiple linear regression [49], as shown in Equation (1) below:

(1)$Vol\left({\mathrm{h}\mathrm{m}}^3\right)={\displaystyle \frac{Depth}{5.61\times {10}^{-4}}}-Area (r=0.985; R^2=0.93;p\le 0.001),$

where “r” is the Pearson correlation coefficient, “R²” is the determination coefficient, “Vol” is the prediction of the total salt marshes’ volume in hm³, “Area” is the equivalent area in km² occupied by the water mass, and “Depth” is the mean value of the water depth, in m, corresponding to the water-occupied area. The depth values were obtained through iterative processes using DEMs or satellite images.

Table 1 shows the depth of the water flooding the Doñana marsh in each of the 17 selected control points.

Next, to confirm the reduction in the flooded surface area in the study region, Figure 6, Figure 7, Figure 8, Figure 9 and Figure 10 display the state of the water surface for the years corresponding to the end of the 2008/2009, 2009/2010, 2010/2011, 2011/2012, and 2012/2013 periods, respectively. Similarly, satellite images for the periods between 2013/2014 and 2019/2020 are also shown to visualize the effect of climate change (see Figure 11, Figure 12, Figure 13, Figure 14, Figure 15, Figure 16 and Figure 17). These images were taken from [51] and processed using Index Stack [52]. This index is a fusion image of Green, Red, Near-Infrared, Red Edge 8, and Shortwave Infrared 1 bands, whereby the vegetation is displayed in green, soil, rocks, and barren land are displayed in blue, and, most importantly, water is represented in magenta.

Based on the above, a study was conducted to explore the potential relationship between precipitation and the presence of water mass in the Doñana salt marshes. Figure 18 illustrates the polynomial relationship between the average monthly value of the normalized difference water index (NDWI), according to Equation (2), and the monthly precipitation from January 2010 to December 2020.

(2)$NDWI={\displaystyle \frac{Green-NIR}{Green+NIR}},$

Similarly, to complement Figure 18, the monthly interconnection between the precipitation and the average NDWI value during the period 2010–2020 was analyzed (Figure 19). This figure shows how, throughout the time series, an increase in precipitation coincides with an increase in the average NDWI value. It can be observed that there is generally a lag between the rise in precipitation and the increase in the average NDWI values, indicating that the flooded area (Figure 20) depends not solely on precipitation.

As a result of the aforementioned observations and to better understand the filling and emptying phenomena in the Doñana marshes, the authors developed a new index, the LAR index (Leiva, Amaro, Ramírez) (see Equation (3) and Figure 21). Values greater than 1 correspond to time periods with no precipitation, coinciding with the onset of the increase in the marsh’s water surface area. Conversely, values lower than 1 indicate time periods where the water surface area decreases while precipitation remains absent.

(3)$LAR index={\displaystyle \frac{Rainfall \left(\mathrm{m}\mathrm{m}\right)-{(NDWI}_{mean}\times 100)}{{NDWI}_{mean}\times 100}}$

After analyzing the LAR index, it became evident that when it presented values greater than 1, the increase in the flooded area must have been due to groundwater from the aquifers in the study area.

Moreover, the Doñana marshlands are currently experiencing a dry period, exacerbated by the exploitation of the Doñana aquifer [53]. This is of significant interest not only due to the precipitation decrease (Figure 22a) and temperature increase (Figure 22b) corresponding to this dry period but also due to excessive use of the aquifer for irrigating agricultural areas and for human consumption. In addition, illegal water extraction has compounded this issue, resulting in short-to-medium-term climate change [53]. Specifically, illegal extraction for irrigated agriculture, particularly for strawberry and blueberry production, has depleted the water reserves and prevented their recovery even in years with abundant rainfall. This has caused the aquifer to deteriorate, from nine stable sectors and seven in poor conditions in 1994 to just five stable sectors and eleven in poor conditions by the end of 2018 [53].

Finally, it is important to highlight that, from 2016 to 2020, as depicted in Figure 22c,d, respectively, the overall trend has resulted in a sharp decrease in precipitation, equivalent to 281 mm (Figure 22c), alongside a 0.5 °C annual average temperature increase (Figure 22d). These results confirm the existence of a dry period and indicate that the temperature increase must be a consequence of higher greenhouse gas emissions from human activities. This can also be seen in Figure 23, where the predictive model of Equation (1) for 2009, 2013, and 2020 is applied.

5. Discussion

First, to determine the existence of significant differences in water depth by control point for the years 2009, 2013, and 2020, as shown in Table 1, a one-way ANOVA was conducted (Table 2). The results indicate that there were no significant differences (p > 0.05) between the years. For this reason, we analyzed the annual clogging rate (marsh sediment deposition) in each of the years specified in Table 1 (see Table 3).

As inferred from Table 4, the ACR varied between 0.75 and 2.86 mm/year, with higher values observed when the ratio between the mean monthly rainfall (MMR) and the mean temperature (MT) of the last year of the analyzed period was lower (see Table 4).

The ongoing dry period trend led to the growth of shrub and bush vegetation, which, over time, retained the soil in the non-flooded marshland [53], increasing the clogging rate.

It is also noteworthy that, annually, as specified by the Guadalquivir River Basin Authority [53] data on the Doñana aquifer, 105 hm³ of water is extracted for irrigation maintenance, exacerbating human-induced climate change as if this extraction trend were to continue alongside global warming in the study area (Figure 22a,b), the marshland completely clogs up (Figure 23).

Applying the model in Figure 23 allowed us to infer the water depth equivalent to the 105 hm³ extracted from the aquifer. For 2009, this volume corresponded to 0.226 m, while for 2013 and 2020 it amounted to 0.223 m and 0.205 m. To determine the proportion of these depths per control point (17 in total), we simply divided them by 17, resulting in 0.0133 m, 0.0131 m, and 0.012 m for 2009, 2013, and 2020, respectively.

Next, we were able to obtain the net average inundated depth by subtracting these proportional depths per control point from the average water depths in the Doñana marshes, as shown in Table 4 (see Table 5). This depth reflects the impact of induced climate change, primarily because the 105 hm³ obtained does not recharge the Doñana marsh itself.

Similarly, considering each of the ACR values from Table 4 as different scenarios (favorable = ACR value from 2009 to 2013; neutral = ACR value from 2009 to 2020; and unfavorable = ACR value from 2013 to 2020), we could determine the number of years remaining until the complete drying of the Doñana marshes (Table 6 and Figure 23).

Additionally, by applying a patented procedure [30], it was possible to annually predict the soil deposited over the Doñana marshes (Figure 24). The simulation was conducted from 2009 to 2025, although the period of 2016–2020 was used as the basis according to the characteristics, precipitation, and temperature of the study area (as it most closely resembles the simulation in the sixth IPCC report [39]).

Using this methodology, which considered the annual aquifer extractions, we could deduce both what happened in the past and what will happen in the future. This approach is a key tool for simulating the effects of human-induced climate change and global warming as well as an indispensable resource for developing recovery and/or improvement plans for the study area.

Finally, it is important to highlight that the Doñana marsh, in relation to the points discussed, does not have sufficient groundwater, and, as such, the disparity between the increase in the inundated area and the water supply from the aquifer will continue to grow. If efficient water resource management measures are not implemented, this will result in the disappearance of the Doñana marsh, its ecosystem, and the climatic benefits that it provides to the surrounding urban areas.

Comparison with Existing Methods

When comparing the method used in this work with other existing ones, one must take into account that the studies conducted to date [8,10,11,12,16,50,53] focus on mapping the situation of the studied wetlands. Although this is important, it should be noted that they leave aside future simulations, through which, alongside big data, the most probable occurrence scenarios can be inferred.

Regardless of the above, the authors used a patented methodology [29], allowing them to obtain soil deposition predictions, which are an extremely useful tool for managing all types of wetlands subjected to both climate change events and human-induced climate change processes.

It is necessary to emphasize that this methodology can be used with data of any type, including data from LIDAR, optical and radar remote sensing, and UAV images, as long as the latter cover a period of time which is broad and significant enough.

6. Conclusions

The Doñana has experienced a significant decline in water availability in recent decades, affecting both its surface and groundwater cycles. The patented procedure employed in this study [30] is a useful tool for helping one discern how much of this process is attributable to human-induced climate change and changes driven by local inhabitants’ resource usage and exploitation, including deforestation, traditional field and crop replacement in watershed areas, the hydraulic management of river channels and flows, extensive modifications to alluvial marshlands, and excessive groundwater extraction in certain areas.

From this perspective, a detailed assessment of the impact of recent and future climatic variations on the Doñana area must necessarily consider the highly anthropized nature of this environment. An analysis of the potential outcomes of combined effects such as decreased or concentrated rainfall and increased evapotranspiration and temperature would be incomplete without accounting for factors like the DEM used, the ground’s shape factor, and the state of the aquifer. Naturally, the use and analysis of satellite images is useful for monitoring what happens with time series data. Without understanding these data series, it would be impossible to reach coherent conclusions.

To mitigate the effects of both global warming and human-induced climate change, it is crucial to apply measures aligned with the conservation of the Doñana, such as, for instance, policies allowing for the sustainable exploitation of the detrital aquifer’s groundwater resources. Such measures would help maintain not only the aquifer’s good condition but also the terrestrial ecosystems dependent on it. Logically, a beneficial measure for maintaining the aquifer would be the closure of all illegal wells and the removal of illegal agricultural crops in the Doñana, efforts which the authorities have been working on for some time.

Although this study was carried out using DEMs from the SRTM, it is undoubtedly recommended, as a future endeavor, to use LIDAR data for the DEM once the time series allows it as this would enhance the results’ precision. Another potential study could compare the salt marshes of the southern Iberian Peninsula with others of the same latitude. In this respect, it is worth noting that the State of Florida in the USA has climatic conditions similar to those of the southern Iberian Peninsula.

Author Contributions

Conceptualization, E.R.-J. and J.-L.A.-M.; methodology, J.L.L.-P., E.R.-J. and J.-L.A.-M.; formal analysis, J.L.L.-P., E.R.-J. and J.-L.A.-M.; investigation, J.L.L.-P., E.R.-J. and J.-L.A.-M.; resources, J.L.L.-P., E.R.-J. and J.-L.A.-M.; writing—original draft preparation, J.L.L.-P., E.R.-J. and J.-L.A.-M.; writing—review and editing, J.L.L.-P., E.R.-J. and J.-L.A.-M.; supervision, E.R.-J. and J.-L.A.-M.; project administration, E.R.-J. and J.-L.A.-M. All authors have read and agreed to the published version of the manuscript.

Institutional Review Board Statement

Not applicable.

Informed Consent Statement

Not applicable.

Data Availability Statement

The original contributions presented in the study are included in the article; further inquiries can be directed to the corresponding author.

Conflicts of Interest

The authors declare no conflicts of interest.

Footnotes

Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content.

Figures and Tables

Figure 1. Study area, own elaboration from [34].

Figure 2. Methodology followed in this research.

Figure 3. Control point distribution. Coordinates in km. Source: own elaboration from [34].

Figure 4. Mean yearly ombrothermic diagram (2006–2020) of the study area. Source: own elaboration from [43].

Figure 5. Land cover changes between 2016 and 2020. Coordinates in km. Source: own elaboration from [49].

Figure 6. Doñana marsh flooding cycle (2008/2009). Own elaboration from [51].

Figure 7. Doñana marsh flooding cycle (2009/2010). Own elaboration from [51].

Figure 8. Doñana marsh flooding cycle (2010/2011). Own elaboration from [51].

Figure 9. Doñana marsh flooding cycle (2011/2012). Own elaboration from [51].

Figure 10. Doñana marsh flooding cycle (2012/2013). Own elaboration from [51].

Figure 11. Doñana marsh flooding cycle (2013/2014). Own elaboration from [51].

Figure 12. Doñana marsh flooding cycle (2014/2015). Own elaboration from [51].

Figure 13. Doñana marsh flooding cycle (2015/2016). Own elaboration from [51].

Figure 14. Doñana marsh flooding cycle (2016/2017). Own elaboration from [51].

Figure 15. Doñana marsh flooding cycle (2017/2018). Own elaboration from [51].

Figure 16. Doñana marsh flooding cycle (2018/2019). Own elaboration from [51].

Figure 17. Doñana marsh flooding cycle (2019/2020). Own elaboration from [51].

Figure 18. Relationship between the monthly average NDWI and the monthly precipitation from 2010 to 2020.

Figure 19. Time series (2010–2020) of monthly average NDWI and monthly precipitation.

Figure 20. Time series (2010–2020) of monthly flooded area and monthly precipitation.

Figure 21. Time series (2010–2020) of the LAR index. Note: the horizontal red line represents the baseline value.

Figure 22. Th evolution of the climatic variables in the study area: (a) precipitation and (b) temperatures from 2006 to 2100 according to [38,39], respectively, and (c) annual average precipitation and (d) temperature from 2016 to 2020.

Figure 23. Predictive model of Equation (1) applied to (a) 2009, (b) 2013, and (c) 2020.

Figure 24. Prediction of soil deposition in the Doñana marshes from 2009 to 2025.

Table 1

Flooded depth by control point in 2009, 2013, and 2020.

Control Point	Water Depth (m) (2009)	Water Depth (m) (2013)	Water Depth (m) (2020)
1	0.005	0.005	0.004
2	0.008	0.008	0.007
3	0.017	0.017	0.015
4	0.032	0.031	0.027
5	0.045	0.044	0.038
6	0.081	0.080	0.069
7	0.111	0.109	0.094
8	0.137	0.135	0.117
9	0.164	0.161	0.139
10	0.191	0.187	0.162
11	0.212	0.208	0.180
12	0.231	0.227	0.196
13	0.249	0.244	0.211
14	0.265	0.260	0.225
15	0.278	0.272	0.236
16	0.296	0.290	0.252
17	0.316	0.310	0.269
Mean value (m)	0.155	0.152	0.132

Table 2

Summary of the analysis of variance (ANOVA) for the estimated depth variable (SS = sum of squares; df = degrees of freedom; MS = mean squared; F = value of the test statistic; and p = significance).

	SS	df	MS	F	p
Treatment (between groups)	0.005434	2	0.002717	0.26	>0.05
Error	0.510406
Total	0.515840

Table 3

Annual clogging rate (ACR) based on the mean water depth for 2009, 2013, and 2020.

	Year2009	Year 2013	Year 2020
Mean depth (m)	0.155	0.152	0.132
Time period	2009–2013	2009–2020	2013–2020
Length (years)	4	11	7
Average depth decrease (m)	0.003	0.023	0.020
ACR (mm/year)	0.75	2.10	2.86

Table 4

Mean temperatures, precipitations, MMR, MT, and MMR/MT for 2009, 2013, and 2020.

Year	Rainfall (mm)	MMR (mm)	MT (°C)	MMR/MT
2009	609	51	18.7	2.72
2013	558	46	18.3	2.51
2020	420	33	18.8	1.86

Table 5

Net average water depth due to induced climate change for 2009, 2013, and 2020.

	Year2009	Year 2013	Year 2020
Mean depth (m)	0.155	0.152	0.132
Mean depth by control points (m)	0.0133	0.0131	0.012
Mean net depth	0.1417	0.1389	0.1200

Table 6

Years until the complete drying of the marsh according to the three ACR scenarios based on 2009, 2013, and 2020.

	Year2009	Year 2013	Year 2020
Mean net depth (m)	0.1417	0.1389	0.1200
Favorable ACR (m/year)	0.00075
Years (Mean net depth/Favorable ACR)	188.93	185.20	160.00
Neutral ACR (m/year)	0.00210
Years (Mean net depth/Neutral ACR)	67.48	66.14	57.14
Unfavorable ACR (m/year)	0.00286
Years (Mean net depth/Unfavorable ACR)	49.55	48.57	41.96

References

1. Quintero-Uribe, L.C.; Navarro, L.M.; Pereira, H.M.; Fernández, N. Participatory scenarios for restoring European landscapes show a plurality of nature values. Ecography; 2022; 2022, e06292. [DOI: https://dx.doi.org/10.1111/ecog.06292]

2. Meacham, M.; Norström, A.V.; Peterson, G.D.; Andersson, E.; Bennett, E.M.; Biggs, R.; Crouzat, E.; Cord, A.F.; Enfors, E.; Felipe-Lucia, M.R. et al. Advancing research on ecosystem service bundles for comparative assessments and synthesis. Ecosyst. People; 2022; 18, pp. 99-111. [DOI: https://dx.doi.org/10.1080/26395916.2022.2032356]

3. Ronkin, V.; Tokarsky, V.; Polchaninova, N.; Atemasov, A.; Koshkina, A.; Savchenko, G. Comparative Assessment of Ecological Plasticity of the Steppe Marmot Between Ukrainian and Kazakhstan Populations: Challenges of the Man-Induced Environmental Changes. Front. Ecol. Evol.; 2020; 8, 219. [DOI: https://dx.doi.org/10.3389/fevo.2020.00219]

4. Umgiesser, G. The impact of operating the mobile barriers in Venice (MOSE) under climate change. J. Nat. Conserv.; 2020; 54, 125783. [DOI: https://dx.doi.org/10.1016/j.jnc.2019.125783]

5. McMichael, C.; Kothari, U.; McNamara, K.E.; Arnall, A. Spatial and temporal ways of knowing sea level rise: Bringing together multiple perspectives. WIREs Clim. Chang.; 2021; 12, e703. [DOI: https://dx.doi.org/10.1002/wcc.703]

6. Paredes, I.; Ramírez, F.; Aragonés, D.; Bravo, M.Á.; Forero, M.G.; Green, A.J. Ongoing anthropogenic eutrophication of the catchment area threatens the Doñana World Heritage Site (South-west Spain). Wetl. Ecol. Manag.; 2021; 29, pp. 41-65. [DOI: https://dx.doi.org/10.1007/s11273-020-09766-5]

7. Camacho, C.; Negro, J.J.; Elmberg, J.; Fox, A.D.; Nagy, S.; Pain, D.J.; Green, A.J. Groundwater extraction poses extreme threat to Doñana World Heritage Site. Nat. Ecol. Evol.; 2022; 6, pp. 654-655. [DOI: https://dx.doi.org/10.1038/s41559-022-01763-6]

8. Adeli, S.; Salehi, B.; Mahdianpari, M.; Quackenbush, L.J.; Brisco, B.; Tamiminia, H.; Shaw, S. Wetland Monitoring Using SAR Data: A Meta-Analysis and Comprehensive Review. Remote Sens.; 2020; 12, 2190. [DOI: https://dx.doi.org/10.3390/rs12142190]

9. Zhang, W.; Liu, Y.; Tang, W.; Wang, W.; Liu, Z. Assessment of the effects of natural and anthropogenic drivers on extreme flood events in coastal regions. Stoch. Environ. Res. Risk Assess.; 2023; 37, pp. 697-715. [DOI: https://dx.doi.org/10.1007/s00477-022-02306-y]

10. Ramos-Fuertes, A.; Prats, J.; Dolz, J. Balances de agua y calor en la marisma de Doñana. Ing. Agua; 2016; 20, 169. [DOI: https://dx.doi.org/10.4995/ia.2016.4720]

11. León-Palmero, E. Understanding the fluxes of greenhouse gases in reservoirs under the inspiration of Margalef. Limnetica; 2023; 42, pp. 302-328. [DOI: https://dx.doi.org/10.23818/limn.42.22]

12. Rashid, I.; Aneaus, S.; Dar, S.A.; Javed, O.; Khanday, S.A.; Bhat, S.U. A novel GIS-based multicriteria analysis approach for ascertaining the catchment-scale degradation of a Himalayan wetland. Environ. Res.; 2023; 229, 115967. [DOI: https://dx.doi.org/10.1016/j.envres.2023.115967] [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/37086883]

13. Varughese, A.R.; Mathew, V. Climate change, flooding woes and mass exodus of inhabitants: An analysis of Kuttanad wetland ecosystem in Kerala, India. GeoJournal; 2023; 88, pp. 6317-6338. [DOI: https://dx.doi.org/10.1007/s10708-023-10971-x]

14. García-Barrón, L.; Aguilar-Alba, M.; Morales, J.; Sousa, A. Classification of the flood severity of the Guadalquivir River in the Southwest of the Iberian Peninsula during the 13th to 19th centuries. Atmósfera; 2022; 36, pp. 1-21. [DOI: https://dx.doi.org/10.20937/ATM.52953]

15. Ramírez-Juidías, E.; Villavicencio-Valero, K.; Borja, A. Sedimentation and Proposed Algorithms to Detect the Possible Existence of Vegetation and Humidity in the Landing Area of the Mars Exploration Rover-B (Opportunity). Solar System Planets and Exoplanets; IntechOpen: London, UK, 2021.

16. Tiengo, R.; Palácios-Orueta, A.; Uchôa, J.; Gil, A. Remote sensing approaches for land use/land cover change in coastal areas and oceanic islands: An open science-based systematic review. Rev. Gestão Costeira Integr.; 2023; 23, pp. 155-177. [DOI: https://dx.doi.org/10.5894/rgci-n552]

17. Sulaiman, M.; Finnesand, E.; Farmanbar, M.; Belbachir, A.N.; Rong, C. Building Precision: Efficient Encoder–Decoder Networks for Remote Sensing Based on Aerial RGB and LiDAR Data. IEEE Access; 2024; 12, pp. 60329-60346. [DOI: https://dx.doi.org/10.1109/ACCESS.2024.3391416]

18. Mack, S.K.; Lane, R.R.; Deng, J.; Morris, J.T.; Bauer, J.J. Wetland carbon models: Applications for wetland carbon commercialization. Ecol. Model.; 2023; 476, 110228. [DOI: https://dx.doi.org/10.1016/j.ecolmodel.2022.110228]

19. Martins, D.; Alves da Silva, A.; Duarte, J.; Canário, J.; Vieira, G. Changes in Vessel Traffic Disrupt Tidal Flats and Saltmarshes in the Tagus Estuary, Portugal. Estuaries Coasts; 2023; 46, pp. 1141-1156. [DOI: https://dx.doi.org/10.1007/s12237-023-01198-7]

20. Zhao, K.; Coco, G.; Gong, Z.; Darby, S.E.; Lanzoni, S.; Xu, F.; Zhang, K.; Townend, I. A Review on Bank Retreat: Mechanisms, Observations, and Modeling. Rev. Geophys.; 2022; 60, e2021RG000761. [DOI: https://dx.doi.org/10.1029/2021RG000761]

21. Ramírez-Juidías, E. Sedimentation Processes in the Tinto and Odiel Salt Marshes in Huelva, Spain. Sedimentation Engineering; InTech: London, UK, 2018.

22. Ramírez-Juidias, E.; Amaro-Mellado, J.-L. Prediction of Erosive Processes in the Río Piedras and Flecha de El Rompido Salt Marsh (Huelva, Spain) Using Iterative Models and GIS. Limnol. Rev.; 2023; 23, pp. 190-202. [DOI: https://dx.doi.org/10.3390/limnolrev23030012]

23. García-López, S.; Vélez-Nicolás, M.; Martínez-López, J.; Sánchez-Bellón, A.; Pacheco-Orellana, M.J.; Ruiz-Ortiz, V.; Muñoz-Pérez, J.J.; Barbero, L. Using UAV Photogrammetry and Automated Sensors to Assess Aquifer Recharge from a Coastal Wetland. Remote Sens.; 2022; 14, 6185. [DOI: https://dx.doi.org/10.3390/rs14246185]

24. Ramírez-Juidías, E. Uso de imágenes satélites para el estudio de los procesos de erosión en las marismas de Isla Cristina (Huelva). South Fla. J. Dev.; 2021; 2, pp. 4828-4836. [DOI: https://dx.doi.org/10.46932/sfjdv2n3-076]

25. Mitchell, M.; Herman, J.; Hershner, C. Evolution of Tidal Marsh Distribution under Accelerating Sea Level Rise. Wetlands; 2020; 40, pp. 1789-1800. [DOI: https://dx.doi.org/10.1007/s13157-020-01387-1]

26. Flor-Blanco, G.; Alcántara-Carrió, J.; Jackson, D.W.T.; Flor, G.; Flores-Soriano, C. Coastal erosion in NW Spain: Recent patterns under extreme storm wave events. Geomorphology; 2021; 387, 107767. [DOI: https://dx.doi.org/10.1016/j.geomorph.2021.107767]

27. Rios, N.N.; Ginesta Torcivia, C.; Onorato, M.R.; Cortez Taillant, V.; Esper Angillieri, M.Y. Morphometry of a Central Andean basin, San Juan, Argentina: Assessment of conditional factors to the occurrence of flash floods. Andean Geol.; 2024; 51, 194. [DOI: https://dx.doi.org/10.5027/andgeoV51n1-3441]

28. Geremew, B.; Tadesse, T.; Bedadi, B.; Gollany, H.T.; Tesfaye, K.; Aschalew, A. Impact of land use/cover change and slope gradient on soil organic carbon stock in Anjeni watershed, Northwest Ethiopia. Environ. Monit. Assess.; 2023; 195, 971. [DOI: https://dx.doi.org/10.1007/s10661-023-11537-7] [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/37466748]

29. Zhang, X.; Fang, C.; Wang, Y.; Lou, X.; Su, Y.; Huang, D. Review of Effects of Dam Construction on the Ecosystems of River Estuary and Nearby Marine Areas. Sustainability; 2022; 14, 5974. [DOI: https://dx.doi.org/10.3390/su14105974]

30. Ramírez-Juidias, E.; Pozo-Morales, L.; Galán-Ortiz, L. Procedimiento para la Obtención de una Imagen Teledetectada a Partir de Fotografía. Spanish Patent; ES2537783B2, 12 June 2015.

31. EarthData (NASA) Shuttle Radar Topography Mission (SRTM). Available online: https://www.earthdata.nasa.gov/sensors/srtm (accessed on 15 April 2024).

32. Ramírez-Juidías, E.; Víquez-Urraco, F.; Noguero-Hernández, D. Sedimentary processes in the Isla Cristina salt-marshes: Geomorphological changes of landscape. Ocean Coast. Manag.; 2017; 143, pp. 148-153. [DOI: https://dx.doi.org/10.1016/j.ocecoaman.2016.11.007]

33. Feagin, R.A.; Chang, K.-A.; Huff, T.P.; Rodriguez-Iturbe, I.; Kim, J.-Y.; Kaihatu, J.; Leonardi, N.; Fagherazzi, S. An asymmetry in wave scaling drives outsized quantities of coastal wetland erosion. Sci. Adv.; 2023; 9, eadj2602. [DOI: https://dx.doi.org/10.1126/sciadv.adj2602]

34. Centro Nacional de Información Geográfica Download Center. Available online: http://centrodedescargas.cnig.es/CentroDescargas/locale?request_locale=en (accessed on 15 May 2024).

35. Green, A.J.; Guardiola-Albert, C.; Bravo-Utrera, M.Á.; Bustamante, J.; Camacho, A.; Camacho, C.; Contreras-Arribas, E.; Espinar, J.L.; Gil-Gil, T.; Gomez-Mestre, I. et al. Groundwater Abstraction has Caused Extensive Ecological Damage to the Doñana World Heritage Site, Spain. Wetlands; 2024; 44, 20. [DOI: https://dx.doi.org/10.1007/s13157-023-01769-1]

36. Mayoral, E.; Díaz-Martínez, I.; Duveau, J.; Santos, A.; Ramírez, A.R.; Morales, J.A.; Morales, L.A.; Díaz-Delgado, R. Tracking late Pleistocene Neandertals on the Iberian coast. Sci. Rep.; 2021; 11, 4103. [DOI: https://dx.doi.org/10.1038/s41598-021-83413-8] [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/33707474]

37. Recio, J.M.; Díaz-delOlmo, F.; Borja, C.; Cámara, R.; Borja, F. Soil formation and hydrological evolution of Navazo del Toro small-lake ecosystem, Doñana National Park, Andalusia, Spain. Ecol. Noospherol.; 2023; 34, pp. 3-13. [DOI: https://dx.doi.org/10.15421/032301] [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/24593738]

38. Rodríguez Ramírez, A.; Yáñez Camacho, C.; Gascó, C.; Clemente Salas, L.; Antón, M.P. Natural and antropic sedimentation rates in the salt marsh of the Doñana National Park: Implications for their handling and conservation. Cuaternario Geomorfol.; 2005; 19, pp. 37-48.

39. Junta de Andalucía. Análisis de la Evolución Futura Bajo Escenarios de Cambio Climático de las Variables Climáticas y de las Variables Derivadas; Junta de Andalucía: Seville, Spain, 2022.

40. Confederación Hidrográfica del GuadalquivirMinisterio para la Transmisión Ecológica y el Reto Demográfico (Gobierno de España). Informe de Estado de los Acuíferos del Entorno de Doñana. Año Hidrológico 2022–2023; Ministerio para la Transición Ecológica y el Reto Demográfico: Seville, Spain, 2024.

41. Abril, J.-M.; Periáñez, R.; Escacena, J.-L. Modeling tides and tsunami propagation in the former Gulf of Tartessos, as a tool for Archaeological Science. J. Archaeol. Sci.; 2013; 40, pp. 4499-4508. [DOI: https://dx.doi.org/10.1016/j.jas.2013.06.030]

42. Instituto Geográfico Nacional PNOA-LiDAR. Available online: https://pnoa.ign.es/web/portal/pnoa-lidar/presentacion (accessed on 10 June 2023).

43. Matgen, P.; Schumann, G.; Henry, J.-B.; Hoffmann, L.; Pfister, L. Integration of SAR-derived river inundation areas, high-precision topographic data and a river flow model toward near real-time flood management. Int. J. Appl. Earth Obs. Geoinf.; 2007; 9, pp. 247-263. [DOI: https://dx.doi.org/10.1016/j.jag.2006.03.003]

44. Ramírez-Juidias, E.; Madueño-Luna, A.; Madueño-Luna, J.M.; López-Gordillo, M.C.; Leiva-Piedra, J.L. Applying Remote Sensing Methods to Estimate Alterations in Land Cover Change and Degradation in the Desert Regions of the Southeast Iberian Peninsula. Remote Sens.; 2023; 15, 3984. [DOI: https://dx.doi.org/10.3390/rs15163984]

45. Climate Research Group (University of Seville, S. Global Climate Monitor). Available online: http://www.globalclimatemonitor.org/# (accessed on 1 May 2024).

46. Phinzi, K.; Ngetar, N.S. The assessment of water-borne erosion at catchment level using GIS-based RUSLE and remote sensing: A review. Int. Soil Water Conserv. Res.; 2019; 7, pp. 27-46. [DOI: https://dx.doi.org/10.1016/j.iswcr.2018.12.002]

47. Romero Pichardo, F. GIS applied to the evaluation of the influence of changes in vegetable coverage on the erosive processes in the basins of the Rocina and Partido streams (Huelva-Spain). Mapping; 2019; 195, pp. 44-54.

48. Ojeda Rivera, J.F.; del Moral Ituarte, L. Water perceptions and management patterns over the different stages of Doñana’s conformation. Investig. Geográficas; 2004; 35, pp. 25-44. [DOI: https://dx.doi.org/10.14198/INGEO2004.35.06]

49. The European Space Agency. Available online: www.esa.int (accessed on 19 July 2024).

50. Sousa, A.; García-Murillo, P.; Sahin, S.; Morales, J.; García-Barrón, L. Wetland place names as indicators of manifestations of recent climate change in SW Spain (Doñana Natural Park). Clim. Chang.; 2010; 100, pp. 525-557. [DOI: https://dx.doi.org/10.1007/s10584-009-9794-9]

51. EOS Data Analytics LandViewer. Available online: https://eos.com/landviewer/ (accessed on 23 June 2024).

52. EOS Data Analytics Index Stack (NDVI, NDWI, NDSI). Available online: https://eos.com/make-an-analysis/index-stack/ (accessed on 23 June 2024).

53. WWF. Doñana, un Acuífero en Alerta Roja. Análisis del “Informe del Estado de los Acuíferos del Entorno de Doñana. Año Hidrológico 2017–2018” de la Confederación Hidrográfica del Guadalquivir; WWF: Seville, Spain, 2019.