Αξιολόγηση και διαχείριση των εξαρτώμενων από τα υπόγεια ύδατα οικοσυστημάτων (GDEs) και οικοσυστημικών υπηρεσιών με την συνδυαστική χρήση μεθοδολογιών χαρτογράφησης και του υπόγειου υδατικού αποτυπώματος (GWF)

Abstract

The present Doctoral Dissertation focuses on the development of tools for the monitoring and management of Groundwater Dependent Ecosystems (GDEs) and the related ecosystem services. Specifically, the Dissertation focuses mainly on the GDEs mapping and on the use of the Groundwater Footprint (GWF) index, as a means of evaluating and managing ecosystem services that rely on groundwater. The literature review conducted in this Dissertation highlighted the increasing recognition of GDEs as critical components of ecological integrity and groundwater sustainability. Although current legislation rarely provides explicit protection frameworks for GDEs, indirect safeguarding is achieved through environmental policies for groundwater status. The main knowledge gaps identified in the literature are summarized as follows: (A) The combined use of remote sensing data and Geographic Information Systems (GIS), integrated into a spatial Multi-Criteria Analysis (MCA), has been increasingly applied in recent years for GDEs mapping. However, the parameters employed as input data in the MCA vary considerably across studies, indicating the need for further investigation, and (B) Regarding the GWF index, the literature highlights the need to identify the relationships between the ecosystem services associated with the GWF and land uses -an important pressure factor affecting their availability, so that the index can be utilized as an effective tool for policy-making. These gaps demonstrate the need for integrated tools that: (a) produce spatially explicit and reproducible GDEs maps, (b) link groundwater-dependent ecosystem services with land uses, and (c) embed these insights into a decision-support framework grounded in the robust groundwater sustainability indicator GWF. Following the literature review, a methodological framework for mapping GDEs through MCA was developed, consisting of the following steps: (a) selection of criteria related to GDEs presence, (b) assessment of criteria correlation, (c) determination of the final set of criteria by eliminating those that are highly correlated with others, (d) min–max normalization of the values of the selected criteria, (e) application of objective weighting methods for the criteria, (f) identification of potential GDEs zones (pGDEz) through a weighted summation model, and (g) validation of the methodology using known spring locations within the study area. The assessment of criteria correlation (step b) is performed to address the problem of the large number and diversity of input data, ensuring the selection of an appropriate final set of MCA parameters and avoiding the inclusion of overlapping information in the model. Furthermore, four of the most commonly encountered objective weighting methods in the literature are applied (step e): the equal weights method, the standard deviation method, the Shannon entropy method, and the CRITIC method. Subsequently, known spring locations are used to evaluate the representativeness of these weighting methods in GDEs mapping (step g). The proposed methodology was applied and validated in the Chania Plain (Crete, Greece). Chania plain is located in the western part of the Region of Crete and is characterized by extended agricultural areas, covering almost 80 % of the total area according to Copernicus Corine Land Cover 2018. The climate of the area is Mediterranean, with mild and rainy winters and warm, dry summers. Numerous springs are met in the area, with Agya spring considered as an ecosystem of great importance providing water supply and irrigation services. The network of streams in the plain is quite complex, with Keritis and Tavronitis being the dominant rivers. It is worth noting that the proposed mapping methodology have been applied in Chania Plain for the years 2017 and 2022 that are selected as typical of the wet and dry conditions, respectively, during the last decade in the area. The 18 criteria identified as predictors of GDEs presence -classified in four categories: conventional GIS layers (e.g. Geology-G); topographic parameters (e.g. Topographic Wetness Index-TWI); multispectral indices (e.g. Normalized Difference Vegetation Index-NDVI); climate variables (e.g. Aridity Index-AI)- are mainly deriving from open-source datasets or processed through GIS tools. The selection of the final set of criteria of the multi-criteria (MCA) model is based on the correlation assessment. In alignment with existing literature, significant correlations (≥0.5) were identified among specific criteria pairs, leading to the exclusion of highly correlated parameters in the final model configuration. Following the criteria correlation assessment, the initial set of 18 criteria/input data was reduced to 11 criteria. As the selected criteria are expressed in different measurement units or scales, a min–max normalization procedure has been applied. Additionally, four of the most common objective weighting techniques —Mean Weight (MW), Standard Deviation (SD), entropy, and Criteria Importance Through Intercriteria Correlation (CRITIC)— have been adopted in order to evaluate the impact of these weighting methods on the delineation of pGDEz. Geology (G) consistently emerged as the predominant factor influencing GDEs presence at an area, followed by the criterion Aridity Index (AI). Moreover, entropy, SD, and CRITIC methods rank flow accumulation (Fa) as the least important criterion, with the lowest relative weight. The resulting maps showcase a prevalence of low to moderate potential GDEs zones in Chania Plain, stable across the weighting methods. Very high pGDEz % coverage for the year 2017 ranging from 10.3 % to 17.3 % according to the weighting method used, while for the year 2022 the relative percentage ranges from 10.3 % to 20.6 %. Comparing the areal extent of pGDEz of the wet and the dry year, the MW method does not identify variance on very high pGDEz; on the contrary, Entropy, SD and CRITIC methods identify an increase of 1.2–4.2 % -depending on the weighing method-in the very high pGDEz of the dry year 2022 compared to the wet year 2017. The validation of the developed model has been carried out by comparing the resulting potential GDEs zone (pGDEz) with spring locations. Based on the results, a high level of model reliability is revealed for all the weighting methods reviewed, as in all cases, none of the springs are located in very low pGDEz and only one is assigned to low pGDEz. However, CRITIC method concludes with the most reliable results with the higher average percentage of springs locations on high and higher potential zones for the two examined years (85 % in year 2017 and 69 % in year 2022). Subsequently, the GWF index is assessed as an indicator for monitoring and evaluating groundwater management in a region, considering the preservation of GDEs and the related ecosystem services. The index focuses on the ecosystem services ‘abstracted water – C’ and ‘groundwater contribution to the environmental flow – E’. Since GWF index does not provide sufficient information on the origin of pressures on ecosystem services or the extent to which these pressures can be mitigated, a methodology was developed to identify the linkages between groundwater-dependent ecosystem services (C and E) and land use/land cover (LULC). The methodology was based on an expert-based survey, in which experts were asked to quantify, using a 3-point Likert scale: (a) the impact of land use/land cover (LULC) on the supply of services C and E, and (b) the importance of services C and E for the maintenance of LULC types. The corresponding questionnaire was completed by 26 experts, primarily from the academic and research community. Τhe majority of experts has indicated C as of highly importance for continuous urban fabric (76.9% of experts), industrial or commercial units (61.5%), green urban areas (88.5%). Discontinuous urban fabric and sport and leisure facilities, i.e., artificial surfaces either vegetated or bare, seem to be considered less water demanding, as in these cases half of the responders have indicated C as moderately important for these specific LULC. Regarding the agricultural areas, C is considered very important for permanently irrigated land (100%), rice fields (84.6%), vineyards (61.5%), fruit trees and berry plantations (80.8%), annual crops associated with permanent crops (73.1%) and complex cultivation patterns (69.2%). On the contrary, for the case of olive trees that are either rainfed or irrigated, more than half of experts (53.8%) have concluded that C is moderately important. C is also considered of moderate importance for LULC classes that are not totally covered by agricultural land, i.e., land principally occupied by agriculture with significant areas of natural vegetation (57.7%) and agro-forestry areas (69.2%). For the case of forest and seminatural areas, wetlands and water bodies, the majority of the experts have indicated no dependance on C, as expected. Almost for most artificial surfaces, E is considered of no to slight importance based on expert opinion, since it is common sense that urban water needs are covered by abstracted water (C). On the contrary, E is considered important for agricultural areas and forests, indicating that the prosperity of ecosystems and communities in these areas are considered to rely on the river’s health. Specifically, for irrigated areas (e.g., permanently irrigated lands and rice fields), more than 70% of the respondents have assigned a score 3 (i.e., very important) regarding E. Regarding E importance for wetlands, expert answers do not lead to a clear consensus, as the most popular answers gather rather low preference rates. Indicatively, coastal wetlands are not considered dependent on E based on 38.5% of experts, while E is considered of moderate importance for inland marshes as stated by 42.3% of experts. The non-consensus may be induced by the fact that even wetlands of the same category (e.g., marshes) may depend on different water sources (e.g., precipitation, groundwater, or surface water). For the cases of broad-leaved, coniferous, and mixed forests, E is considered of high importance, probably due to the dependence of the riparian forests on river flow. Experts have identified the high importance of E for the maintenance of water bodies and courses as well as estuaries, since E ensures river health. Experts identified a negative impact of continuous urban fabric and industrial/commercial units due to the advanced water needs, but also by the pressure on groundwater quality. The impact of discontinuous urban fabric on C is characterized as neutral by the majority of experts as expected, since the built-up areas density is lower than in continuous urban fabric, and therefore, water needs are lower. Additionally, the existence of vegetated areas and bare surfaces enables groundwater recharge. Airports and sport and leisure facilities are assigned neutral impact on C, despite the large water needs, mainly for non-potable water, probably due to the expected existence of vegetated areas and bare surfaces. For the case of agricultural areas, LULC that imply irrigation needs (e.g., permanently irrigated land, rice fields and vineyards) have been described by the overwhelming majority of experts, as having a negative impact on C availability. For the case of olive trees, expert opinion is almost equally divided between neutral impact (50%) and negative impact (42%), since groves are either rainfed or irrigated. For all sub-classes of forest and semi-natural areas, expert opinion converges to positive impact on C availability, as no pressure on groundwater resources is observed. On the contrary, forest and semi-natural areas enable groundwater recharge. For the case of wetlands and water bodies, a positive impact is identified, probably due to their eventual contribution to groundwater recharge. As proposed in the literature, the results of the survey presented as correlation matrices between LULC types and services C and E can be further utilized for mapping these ecosystem services. Finally, an integrated framework was proposed for the evaluation of ecosystem services C and E and the management of related ecosystems, based on the combined use of the proposed tools and the GWF index. This 6-steps framework highlights the synergy between the proposed tools and the GWF index, which can contribute to the effective development of land-use change scenarios and their comparative assessment in terms of maintaining or enhancing priority ecosystem services. According to this methodological framework, Step 1 concerns the mapping of GDEs and, more specifically, the identification of potential GDE zones (pGDEz), reflecting the current state of the ecosystem. Step 2 involves the assessment of the current condition of the ecosystem services C and E through the estimation of the GWF index under existing land-use patterns. This estimation requires the development of a groundwater flow model that accounts for the morphological and hydrological characteristics of the area, existing infrastructure, land uses, and anthropogenic pressures exerted on the groundwater system. The GWF value under current conditions serves as a baseline against which the deviations introduced by future scenarios can be evaluated. Steps 3 and 4 relate to the formulation of alternative land-use scenarios. This process is grounded in the information obtained from the expert survey regarding the interactions between LULC and ecosystem services. Scenario formulation is based on (a) identifying the priority ecosystem services according to future management objectives (Step 3), and (b) defining the alternative land-use configurations that could improve ecosystem management regarding the maintenance of the desired level of the priority ecosystem service (Step 4). Step 5 refers to the estimation of the GWF index for each scenario, serving as the key quantitative tool for evaluating the management of services C and E. The results produced at this step support communication with stakeholders, decision-makers, and the public and thereby inform Step 6, which concerns the final management decision. It is worth noting that using the information derived from the expert survey can substantially reduce the number of scenarios that must be evaluated, thus decreasing the computational burden associated with running the groundwater flow model. Overall, the integration of GDEs mapping, expert-based ecosystem service - LULC linkages and the GWF index provides a coherent and operational framework that can support groundwater management and foster the long-term resilience of groundwater-dependent ecosystems.

Αντικείμενο της παρούσας Διδακτορικής Διατριβής αποτελεί η ανάπτυξη εργαλείων για την παρακολούθηση και διαχείριση των Οικοσυστημάτων που Εξαρτώνται από τα Υπόγεια Ύδατα (Groundwater Dependent Ecosystems – GDEs) και των σχετικών οικοσυστημικών υπηρεσιών, εστιάζοντας κυρίως στην χαρτογράφησή τους και στην εφαρμογή του δείκτη Υπόγειο Υδατικό Αποτύπωμα (Groundwater Footprint – GWF) για τη διαχείριση των οικοσυστημικών υπηρεσιών που παρέχονται από τα GDEs. Από την ανασκόπηση της βιβλιογραφίας διαπιστώνεται ότι η συνδυασμένη χρήση δεδομένων τηλεπισκόπησης και Γεωγραφικών Συστημάτων Πληροφοριών (Geographic Information Systems – GIS) και η ενσωμάτωση τους σε μία χωρική Πολυκριτηριακή Ανάλυση (Multi-Criteria Analysis – MCA) αποτελεί μία μεθοδολογία που τα τελευταία χρόνια εφαρμόζεται στη χαρτογράφηση των GDEs παρουσιάζοντας αρκετά πλεονεκτήματα. Ωστόσο, οι παράμετροι που χρησιμοποιούνται ως δεδομένα εισόδου στη συγκεκριμένη μεθοδολογία παρουσιάζουν μεγάλη ποικιλομορφία γεγονός που υποδεικνύει την ανάγκη για περαιτέρω διερεύνηση. Όσον αφορά στον δείκτη GWF, από τη μελέτη της βιβλιογραφίας προκύπτει η ανάγκη για εντοπισμό των σχέσεων μεταξύ των οικοσυστημικών υπηρεσιών που σχετίζονται με τον δείκτη GWF και των χρήσεων γης έτσι ώστε ο δείκτης να μπορέσει να αξιοποιηθεί ως ένα αποτελεσματικό εργαλείο για τη χάραξη πολιτικών. Σε συνέχεια της βιβλιογραφικής ανασκόπησης αναπτύσσεται ένα μεθοδολογικό πλαίσιο για τη χαρτογράφηση των GDEs μέσω πολυκριτηριακής ανάλυσης (MCA) το οποίο αποτελείται από τα παρακάτω βήματα: α) επιλογή κριτηρίων που σχετίζονται με την παρουσία GDEs, β) εκτίμηση της συσχέτισης κριτηρίων, γ) προσδιορισμός του τελικού σετ κριτηρίων, εξαλείφοντας τα κριτήρια που συσχετίζονται σε μεγάλο βαθμό με άλλα κριτήρια, δ) κανονικοποίηση ελαχίστου-μεγίστου (min-max normalization) των τιμών των επιλεγμένων κριτηρίων, ε) εφαρμογή μεθόδων αντικειμενικής στάθμισης των κριτηρίων, στ) προσδιορισμός των pGDEz μέσω μοντέλου το οποίο βασίζεται στη μέθοδο βαρύνουσας άθροισης (weighted summation) και ζ) επικύρωση της μεθοδολογίας μέσω της αξιοποίησης γνωστών θέσεων πηγών στην περιοχή εφαρμογής. Επισημαίνεται ότι η εκτίμηση της συσχέτισης των κριτηρίων (βήμα β) πραγματοποιείται προκειμένου να αντιμετωπιστεί το πρόβλημα του μεγάλου πλήθους και της διαφοροποίησης των δεδομένων εισόδου, ώστε να επιλεγεί το κατάλληλο τελικό σετ παραμέτρων πολυκριτηριακής ανάλυσης και να αποφευχθεί η εισαγωγή αλληλεπικαλυπτόμενης πληροφορίας στο μοντέλο. Επισημαίνεται επίσης ότι εφαρμόζονται τέσσερις εκ των πιο συχνά απαντώμενων –σύμφωνα με τη βιβλιογραφία– αντικειμενικών μεθόδων στάθμισης (βήμα ε): η μέθοδος των ισοβαρών κριτηρίων, η μέθοδος της τυπικής απόκλισης, η μέθοδος της εντροπίας Shannon και η μέθοδος CRITIC. Στη συνέχεια αξιοποιούνται γνωστές θέσεις πηγών για τη διερεύνηση της αντιπροσωπευτικότητας των υπό αξιολόγηση αντικειμενικών μεθόδων στάθμισης στη χαρτογράφηση GDEs (βήμα ζ). Η προτεινόμενη μεθοδολογία εφαρμόστηκε και επικυρώθηκε στην περιοχή του Κάμπου Χανίων. Κατά την εφαρμογή το αρχικό σετ των 18 κριτηρίων/δεδομένων εισόδου μειώθηκε σε 11 κριτήρια/δεδομένα εισόδου. Η επικύρωση των αποτελεσμάτων πραγματοποιήθηκε και για τις τέσσερις μεθόδους στάθμισης μέσω της αξιολόγησης της κατηγορίας των δυνητικών ζωνών GDEs που προέκυψε από την εφαρμογή της μεθοδολογίας σε γνωστές θέσεις πηγών· στις θέσεις των πηγών αναμένονται ζώνες υψηλού δυναμικού GDEs. Για όλες τις μεθόδους αντικειμενικής στάθμισης που εξετάστηκαν, το ποσοστό συμφωνίας μεταξύ θέσεων πηγών και ζωνών υψηλού δυναμικού GDEs κρίθηκε ικανοποιητικό με βάση τη βιβλιογραφία· ωστόσο η μέθοδος CRITIC αναδείχθηκε ως η πιο αντιπροσωπευτική μέθοδος μεταξύ των τεσσάρων μεθόδων στάθμισης (υψηλότερο ποσοστό συμφωνίας). Η αντιπροσωπευτικότητα της μεθόδου CRITIC επικυρώθηκε επίσης μέσω μετρήσεων της στάθμης του υπόγειου νερού· τα ρηχά υπόγεια νερά (shallow groundwater) υποδεικνύουν μεγαλύτερη πιθανότητα ύπαρξης GDEs. Ακολούθως ο δείκτης GWF μελετάται ως ένας δείκτης παρακολούθησης και αξιολόγησης της διαχείρισης των υπόγειων υδάτων μιας περιοχής, λαμβάνοντας υπόψη τη διατήρηση των GDEs και των σχετιζόμενων οικοσυστημικών υπηρεσιών. Ο δείκτης εστιάζει στις οικοσυστημικές υπηρεσίες ¬«αντλούμενο νερό – C» και «συνεισφορά των υπόγειων υδάτων στην περιβαλλοντική ροή – E». Καθώς ο δείκτης GWF δεν προσφέρει επαρκείς πληροφορίες σχετικά με την προέλευση της πίεσης στις οικοσυστημικές υπηρεσίες και τον βαθμό στον οποίο αυτή μπορεί να περιοριστεί, στο πλαίσιο της Διδακτορικής Διατριβής αναπτύχθηκε μια μεθοδολογία για τον προσδιορισμό των δεσμών των –εξαρτώμενων από τα υπόγεια ύδατα– οικοσυστημικών υπηρεσιών που σχετίζονται με τον υπολογισμό του δείκτη GWF (C και E) και των χρήσεων/κάλυψης γης (LULC). Η μεθοδολογία βασίστηκε στην έρευνα μέσω εμπειρογνωμόνων (expert-based), οι οποίοι κλήθηκαν να ποσοτικοποιήσουν με τη χρήση κλίμακας Likert: (α) την επίδραση των χρήσεων/κάλυψης γης (LULC) στην προσφορά των C και Ε και (β) τη σημαντικότητα των οικοσυστημικών υπηρεσιών C και E για τη διατήρηση των χρήσεων/κάλυψης γης (LULC). Το σχετικό ερωτηματολόγιο απαντήθηκε από 26 εμπειρογνώμονες οι οποίοι προέρχονται κυρίως από την ακαδημαϊκή και την ερευνητική κοινότητα. Τα αποτελέσματα της έρευνας αποτυπώθηκαν σε μήτρες συσχέτισης των χρήσεων/κάλυψης γης και των υπηρεσιών C και Ε, οι οποίες μπορούν να αξιοποιηθούν για τη χαρτογράφηση αυτών των οικοσυστημικών υπηρεσιών. Τέλος, προτείνεται ένα ολοκληρωμένο πλαίσιο αξιολόγησης των οικοσυστημικών υπηρεσιών C και E και διαχείρισης των σχετιζόμενων οικοσυστημάτων, το οποίο στηρίζεται στη συνδυαστική χρήση των προτεινόμενων εργαλείων και του δείκτη GWF. Το προτεινόμενο πλαίσιο αναδεικνύει τη συνέργεια μεταξύ των εργαλείων και του δείκτη, η οποία μπορεί να συμβάλει στην αποτελεσματική ανάπτυξη σεναρίων αλλαγής χρήσεων γης και στη συγκριτική αξιολόγησή τους, όσον αφορά τη διατήρηση ή την ενίσχυση των οικοσυστημικών υπηρεσιών προτεραιότητας.