*2.2. Methodology*

In this work, the definition of the Prolonged Drought and Scarcity indicators for the Portuguese part of Minho and Lima river basins was carried out according to the methodology defined in the current version of the Special Drought Plan of the Spanish part of the Miño-Sil River Basin District (2018 PES-MS) [20], in agreemen<sup>t</sup> with the technical instruction provided by the Ministry for the Ecological Transition of Spain [28].

A prolonged drought is a natural, persistent and intense situation of reduction of precipitation, produced by unusual circumstances, with influence on the runoff. The Prolonged Drought Indicator (PDI) should identify, temporally and territorially, runoff reduction by natural causes, independently of human water resources management. Therefore, the aim of the PDIs is to establish the threshold for compliance with the environmental flow regime defined in the RBMPs and to limit the occurrence of situations of temporary deterioration of water bodies quality only to prolonged drought natural phenomena situations (and not to scarcity situations), as set out in the WFD [6,28]. According to Article 18 of the Spanish Hydrological Planning Regulation [29], a prolonged drought situation allows the justified reduction of the environmental flows of water bodies established in the RBMPs.

A scarcity situation is defined as a temporal problem of lack of resources to meet the water demands associated with the different socioeconomic uses of water. Thus, the Scarcity Indicator (SI) is based on the relationship between the availability of resources and water demands, identifying the inability of the resources to meet the demands. Consequently, it serves as an instrument of assistance in decision making related to the managemen<sup>t</sup> of water resources. The SI is an operational indicator, aiming at the progressive triggering of measures in order to postpone or avoid the occurrence of the most severe stages of scarcity, mitigating their adverse impacts to the several water uses [6,28].

The general methodology used for the definition of each of the two indicators is schematized in Figure 3.

**Figure 3.** General methodology for the definition of both indicators (PDI and SI) (adapted from [20]).

The procedure begins with the selection of the hydrometeorological variables for each defined territorial unit of analysis. For the PDI, two types of variables may be considered: precipitation and runoff in natural regime. For the SI, the variables may be of various types, including runoff, reservoir inflow, reservoir storage volume, snow storage, and groundwater levels. These variables can be represented by their own values or by the corresponding standardized indices (e.g., SPI).

After the selection of the variables, they are rescaled into dimensionless variables varying between 0 and 1. In order for the rescaling of the variables to be performed, both indicators require a reference time series to be considered. The period between October 1980 and September 2012 was considered for all the reference series displayed in the 2018 PES-MS [20]. This 32-year hydrological monthly time series served as a sample to perform a statistical analysis of the data records, identifying the characteristics of the humid, normal and dry periods. Therefore, it allowed us to establish the predominant values/characteristics in the TMUs and enabled the analysis and characterization of later events by comparing it with the values of the reference series.

For both the Prolonged Drought and Scarcity indicators, the rescaling step requires the establishment of monthly thresholds for each of the 12 months of the year. One threshold was defined for PDI, corresponding the occurrence of prolonged drought, and three thresholds were defined for SI, corresponding to different limits of water scarcity stages (pre-alert, alert and emergency).

Subsequently, the rescaled variables were aggregated in a weighted way, producing a single indicator (also entitled as a Status Index, or "Indice de Estado" in Spain). This has a grea<sup>t</sup> advantage relative to the use of other type of indicators, namely simple standard indexes, because it enables to compare the indicator's results among different basins, despite their diverse geographical, climatic, water demands and other specific characteristics [6].

As referred before (in Section 1), the territorial units most suitable for the analysis and managemen<sup>t</sup> of the two situations (prolonged drought and water scarcity) may be different. In the case of the 2018 PES-MS, the territorial units of analysis used for both indicators' definitions were those presented in Figure 2. Similarly, for the work developed and presented here, it was considered that the territorial units used for the Prolonged Drought and Scarcity indicator definitions in the Portuguese parts of the Minho and Lima river basins corresponded to their full country areas (also pictured in Figure 2).

In the Portuguese parts of the Minho and Lima river basins, the PDI and SI were defined for the period from October 1980 to September 2017, taking as reference time series the period from October 1980 to September 2012, such as in the 2018 PES-MS [20]. For the Lima river basin, the SI was defined for the period from October 1993 to September 2017, taking as reference time series the period from October 1993 to September 2012, as will be reported below (in Section 2.2.2).

The indicators obtained to the Portuguese parts of the Minho and Lima river basins were then compared, for each basin, with those obtained in the corresponding Spanish parts of the basins. In this context, the Portuguese parts of the Lima basin were compared with the ones defined to the Limia TMU. For the Minho basin, a similar comparison was made with the Miño-Bajo TMU, as this was the nearest territorial unit adjacent to the Portuguese part (see Figure 2).

The specific methodology used to define the Prolonged Drought and the Scarcity indicators, respectively, in the Portuguese parts of the Minho and Lima river basins are described below (in Sections 2.2.1 and 2.2.2), in more detail.
