1. Determine the spatial unit (SU) and produce a GIS layer that shows their location and coverage of the basin**.**

  2. Determine the time-period (or evaluation period) for the indicator calculation. If the calculation is to go beyond F1, the evaluation period has to be divided into smaller duration of time (termed here as ‘instances’). For example, for an evaluation period of 5 years, each year may be considered as an instance over which to group the events. The test of whether a demand is met or not is conducted within the period represented by the instance.

  3. Determine from the data whether F1 and F2 can be calculated. If information regarding the number of SUs affected by lack of delivery of ecosystem services is available, then F1 can be calculated using the following formula:

    $$ F1 = \ \left( \frac{\text{Number of SUs that did not meet demand at least once}}{\text{Total number of SUs}} \right) \times 100 $$

    If distribution of events where demand is not met is available over the evaluation period is available, then F2 can be calculated by considering over which instances was demand met or not met:

    $$ F2 = \ \left( \frac{\text{Number of instances where demand was not met}}{\text{Total number of instances monitored}} \right) \times 100 $$

  4. Determine from the data if F3 can be calculated and is ‘sharp’ or ‘fuzzy.’ If any information for the ability to meet demand over the instances and magnitude of departure is available, it may be possible to calculate F3. This evaluation procedure is encapsulated within a measure of ‘excursion’ for each instance that is deemed to return a non-compliant value. Excursion for each instance i (Exi) can be calculated as follows:

    a. Services where a univariate ‘sharp’ threshold for non-compliance can be defined:

    Here, an objective value (such as target volume to meet water demand) for that particular instance can be defined, and thus, excursion for each case where demand is not met can be evaluated.

    When target is to not fall short of this objective, excursion can be defined as:

    $$ \text{Ex}_{i} = \ \left( \frac{\text{objective}_{i}}{\text{instance value}_{i}} \right) - 1 $$

    Alternately, when the target is to not exceed the objective, excursion can be defined as:

    $$ \text{Ex}_{i} = \ \left( \frac{\text{instance value}_{i}}{\text{objective}_{i}} \right) - 1 $$

    b. Services where a univariate ‘sharp’ threshold for non-compliance cannot be defined:

    Here, a single objective may be hard to define. We recommend that in these, excursion for each instance i be ranked on a scale of 1 to 10 to correspond with low to high gap between demand and supply. The values can be defined through stakeholder surveys or through tracking and combining a few metrics relevant to the ecosystem service. For n instances among the SUs where objective is not met is then collated and into a normalized sum of excursions (nse) such that:

    $$ nse = \frac{\sum_{i = 0}^{n}\text{Ex}_{i}}{\text{Total number of instances monitored}} $$

    Note that for the normalization process, the total number of instances monitored – whether demand is met or not met – is used. This is done so that the excursions are scaled with respect to all information available about the system and not biased toward instances where demand is not met.

    Finally, F3 is now calculated by scaling nse to a 0-100 scale using the asymptotic function proposed by Saffran et al. (2001):

    $$ F3 = \left( \frac{\text{nse}}{nse + 1} \right) \times \ 100 $$

  5. Combine F1, F2 and F3.
    • If able to only determine F1 (low evidence): $$ESI = 100\ - \ F1$$

    • If able to only determine F1 and F2 (medium evidence): $$ESI\ = \ 100\ - \ \sqrt{F1 \times F2}$$

    • If able to determine all three (high evidence): $$ESI\ = \ 100\ - \ \sqrt{F1 \times F3}$$