The computation methods is presented below for each of the seven sub-indicators: (1) Permanent and Seasonal Surface Water; (2) Reservoirs; (3) Wetlands; (4) Mangroves; (5) Water Quality: Turbidity and Trophic state; (6) River Flow; (7) Groundwater; as well as for the computation of a (8) national summary score based on the integration of all sub-indicators.

Permanent and Seasonal Surface Water

Description of the method used to globally map all surface water:

Data on the spatial and temporal dynamics of naturally occurring surface water has been generated for the entire globe. A Global Surface Water dataset (Pekel et al., 2016) has been produced by the European Commission's Joint Research Centre. The dataset documents different facets of the long term (since 1984 onward) water dynamics at 30x30 meter pixel resolution. The dataset documents permanent and seasonal surface water surfaces. All naturally occurring surface water larger in area than 30x30 meters has been mapped and at this 30-meter grid/pixel spatial resolution satellite imagery is predominantly capturing areas of lakes and wide rivers. The data include land areas that are temporarily inundated such as wetlands and paddy fields. Smaller rivers and waterbodies are not captured as they are too narrow to detect or are masked by forest canopy. The data include individual full-resolution images acquired by the Landsat 5, 7 and 8 satellites. These satellites capture images which are distributed publicly by the United States Geological Survey. Together they provide multispectral imagery at 30x30 meter resolution in six visible, near and shortwave infrared channels, plus thermal imagery at 60x60 meters.

The data includes land surfaces that are under water (e.g. a permanent water area) for all twelve months of a year. It also accounts for seasonal and climactic fluctuations of water, meaning lakes and rivers which freeze for part of the year are captured. Areas of permanent ice, such as glaciers and ice caps as well as permanently snow-covered land areas are not included. Areas of consistent cloud cover inhibit the observation of water surfaces in some areas and in these limited locations optical observations may not be available. A global shoreline mask has been applied to the data to prevent ocean water being included in the freshwater statistics and the methodology for this shoreline mask is published in the journal of operational oceanography (Sayer et al. 2019).

The surface water maps are derived from the analysis of over four million images collected over 36 years which have been individually processed using an expert system classifier. The accuracy of the Global Surface Water map was determined using over 40,000 control points from around the world and across the 36 years. The full validation methodology and results have been published in the scientific journal Nature (cf. Pekel et al., 2016). The validation results show that the water detection expert system produced less than 1% of false water detections, and that less than 5% of water surfaces were missed.

In addition to reporting the temporal changes in the permanent and seasonal water area the SDG 6.6.1 data portal (www.sdg661.app) also documents various water transitions relating to permanent and seasonal surface water - these are changes in water state between two points in time. Data is available for various transitions including new permanent water surfaces (i.e., conversion of a no water place into a permanent water place.); lost permanent water surfaces (i.e., conversion of a permanent water place into a no water place) as well as new and lost seasonal water. These allow monthly water presence or absence data to be captured. It is possible to identify specific months/years in which conditions changed, e.g., the date of filing of a new dam, or the month/year in which a lake disappeared. In addition, data on seasonality are provided, capturing changes resulting from intra and inter-annual variability or resulting from appearance or disappearance of seasonal or permanent water surfaces. The data separates 'permanent' water bodies (those that are present throughout the period of observation) [nominally a year] from 'seasonal' (those that are present for only part of the year).

Calculating the change in surface area of permanent and seasonal surface water:

Data on monthly surface water dynamics are available for a 38-year period, from 1984-onward. Every year new annual data for seasonal and permanent water extent is produced and added to this time series. The SDG indicator 6.6.1. was developed to measure changes in freshwater ecosystems between 2000 and 2030. Since freshwater ecosystems (including surface-water bodies) are dynamic, a long time series of annual data is needed to identify changes that depart significantly from the longer-term mean. Changes in surface-water bodies are therefore measured in five-year intervals relative to a 20-year reference period (2000-2019) and based on the annual aggregation of monthly water occurrence maps derived from a time series of Landsat data. Mathematically the change of spatial extent of permanent and seasonal waters is calculated using equation 1:

$E q u a t i o n 1 : ∆ = \frac{γ - β }{β} \times 100$

Where:

$∆$ – percentage change in spatial extent;

$β$ – the median spatial extent for the baseline reference period (2000-2019);

$γ$ – the median spatial extent for the most recent 5-year reporting period (e.g., 2018-2022).

Equation 1 is applied to measure changes in both permanent waters (cf. water that is observable year-round) and seasonal waters (cf. where water is observed for less than 12 months of the year).

The nature of this formula yields percentage change values as either positive or negative, which helps to indicate how spatial area is changing. On the SDG661 data portal, statistics are displayed using both positive and negative symbols. For interpretation of the statistics, if the value is shown as positive, the statistics represent an area gain while if the value is shown as negative, it represents a loss in surface area.

The use of ‘positive’ and ‘negative’ terminology does not imply a positive or negative state of the water-related ecosystem being monitored. Gain or loss in surface water area can be beneficial or detrimental. The resulting impact of a gain or loss in surface area must be locally contextualized. The percentage change statistic produced represents how the total area of lakes and rivers within a given boundary (e.g., nationally) is changing over time. Percentage change statistics aggregated at a national scale should be interpreted with some degree of caution because these statistics reflect the areas of all the lakes and rivers within a country boundary. For this reason, sub-national statistics are also made available including at basin and sub-basin scales. The statistics produced at these smaller scales reflects area changes to a smaller number of lakes and rivers within a basin or sub-section of a basin, allowing for localized, water body specific, decision making to occur.

Reservoirs

Description of the method used to globally map changes to reservoir surface area:

A global reservoir dynamics dataset has been produced by the European Commission's Joint Research Centre. The dataset documents the long term (since 1984 onward) spatial area dynamics of 8,869 reservoirs at 30x30 meter pixel resolution. The reservoirs dataset represents surface area data on artificial waterbodies including reservoirs formed by dams, flooded areas such as opencast mines and quarries, and water bodies created by hydro-engineering projects such as waterway and harbour construction. The reservoirs dataset is derived from the Global Surface Water Explorer (GSWE) dataset, onto which is applied an expert system classifier designed to separate natural and artificial water bodies. The expert systems classifier is non-parametric to account for uncertainty in data, incorporate image interpretation expertise into the classification process, and uses multiple data sources. The expert system has been developed to delineate natural and artificial water using an evidential reasoning approach; the geographic location and the temporal behaviour of each pixel; and fed with the following datasets:

Global Surface Water Explorer (Pekel et al., 2016): This dataset that maps the location and long term (since 1984 onward) temporal distribution of water surfaces at global scale. The maps show different facets of surface water dynamics and document where and when open water was present on the Earth's surface. The maps include natural (rivers, lakes, coastal margins and wetlands) and artificial water bodies (reservoirs formed by dams, flooded areas such as opencast mines and quarries, flood irrigation areas such as paddy fields, and water bodies created by hydro-engineering projects such as waterway and harbour construction). The complete history of any water surface can be accessed at the pixel scale as temporal profile. These profiles allow for identifying specific months or years during which conditions changed, e.g. the date on which a new dam was created, or the month or year in which a lake disappeared. The GSWE dataset is continuously updated providing consistent global monitoring of open water bodies.
Global Reservoir and Dam Database (Lehner et al, 2011): The Global Reservoir and Dam Database v1.3 is the output of an international effort to collate existing dam and reservoir datasets with the aim of providing a single, geographically explicit and reliable database for the scientific community. The initial version (v1.1) of GRanD contains 6,862 records of reservoirs. The latest version (v1.3) augments v1.1 with an additional 458 reservoirs and associated dams to bring the total number of records to 7320.
Global Digital Surface Model: ALOS World 3D - 30m is a global digital surface model (DSM) dataset with a horizontal resolution of approximately 30 meters (1 arcsec mesh). The dataset is based on the DSM dataset (5-meter mesh version) of the World 3D Topographic Data. More details are available in the dataset documentation here.
Digital Elevation Data (Farr et al, 2004): The Shuttle Radar Topography Mission (SRTM, see Farr et al. 2007) is a digital elevation dataset at 30 meters resolution provided by NASA JPL at a resolution of 1 arc-second.

Calculating the extent to which reservoir area is changing over time:

Data on reservoir areas are available on monthly basis for a 38-year period, from 1984-onward. Every year new annual data of minimum and maximum reservoir water area is produced and added to this time series. To calculate percentage change in reservoir area a long-term baseline period has been defined and to be compared against any subsequent 5-year target period. Mathematically the change of spatial extent of reservoirs is calculated using equation 2:

$E q u a t i o n 2 : ∆ = \frac{γ - β }{β} \times 100$

Where:

$∆$ – percentage change in spatial extent;

$β$ – the median spatial extent for the baseline reference period (2000-2019);

$γ$ – the median spatial extent for the most recent 5-year reporting period (e.g., 2018-2022).

Equation 2 is applied to measure changes in minimum reservoir extent, which is arguably the most critical parameter for reservoir monitoring.

Minimum water extent of reservoirs is the lowest observed (or minimum) surface area of reservoirs in a year (intra-annual measurement). This minimum extent varies from one year to another. The data shows the extent to which the annual minimum surface area of reservoirs has changed compared to a reference period. Change is either gain or loss both shown in both percentage and km² units.

Known limitations and scope for improvements.

The current version of the Global Reservoir Dynamics dataset has the following known limitations:

- Some reservoirs built prior 1984 may be missing;

- Reservoirs smaller than 3 hectares (30 000 square meters) may be missing;

- Branches of reservoirs whose width is smaller than 30 meters may be missing.

- The Global Reservoir Dynamics relies on SRTM/ALOS DEM but new improved DEMs are available (e.g. GLO-30 DEM^[1])

Wetlands

Description of the method used to globally map wetlands:

Inland vegetated wetlands are mapped according to the following definition: “Inland vegetated wetlands include areas of marshes, peatlands, swamps, bogs and fens, the vegetated parts of flood plains as well as rice paddies and flood recession agriculture”. This sub-indicator only measures inland vegetated wetlands and not coastal mangroves (see section 3.5 of this methodology on mangroves). This SDG indicator methodology is used for official reporting of SDG indicator 6.6.1 statistics. A high-resolution global geo-spatial mapping of inland vegetated wetlands has been produced detailing the spatial area of wetlands per country. The data on wetlands has been produced to support countries with monitoring their wetland ecosystems and bridge an existing global data gap. The data production method uses a consistent wetland monitoring mechanism based on satellite Earth Observation data and the global map includes the entire land surface of Earth except for Antarctica and a few small islands. As wetlands tend to be susceptible to high annual variations, multi-annual data was collected to even out potential annual biases and create a robust estimate of wetland area. Data was gathered from 2016, 2017 and 2018 and combined to produce a wetlands area baseline measurement (in km²).

Future annual updates will enable wetlands change statistics to be produced and once available these will be displayed on the SDG 6.6.1 data portal. Predicting wetland area using Earth Observation data relies on four components: stratification, training data, machine learning, and post-processing. The approach uses all available data from the satellites Sentinel-1, Sentinel-2, and Landsat 8 to predict wetland probability. A Digital Elevation Model is used to qualify wetland predictions and a post-processing routine converts the wetland probability map into a map of wetland area. In addition, topographic information from satellite-derived Digital Elevation Models (DEMs) are used. Close to 4 million satellite images amounting to 2.8 petabyte of data were analysed and classified as wetland or non-wetland using an automated machine learning model. Users of the global wetland map should be aware that the map represents a first line rapid assessment of the global distribution of vegetated wetlands. The methodology applied identifies vegetated inland wetlands. This may generate underestimations compared to national statistics which may integrate metrics on surface water and coastal/marine wetlands.

Figure 1. Workflow for mapping global wetland area

Data accuracy for the available wetlands data is approximately 70% and wetland data with 100% accuracy is not feasible at this current time. While it is based on a scientifically sound and robust mapping approach, there will inevitably be inaccuracies in the wetland predictions both in terms of commission and omission errors. Notable commission errors are for instance high-intensive irrigated agriculture parcels being classified as wetlands because they resemble many of the inherent spectral characteristics of wetlands (i.e. high moisture and vegetation presence even in dry season). Omission errors will mainly be attributed to the large diversity of wetlands. It is also worth noting that since the map only considers vegetated wetlands it may generate underestimations compared to national statistics which typically integrate metrics on surface water and coastal/marine wetlands.

Calculating the change in surface area of wetlands per country:

No change in surface area has yet been calculated. However, a baseline surface area has been calculated per country. This methodology uses a 2017 baseline (based on input imagery data from 2016 to 2018 to even out potential annual biases). Going forward, updates to this wetland area datasets will be produced annually. Once the update is produced it will be possible to calculate change of wetland area from the baseline reference period. Using this baseline period, percentage change of spatial extent is calculated using equation 3:

$E q u a t i o n 3 : ∆ = \frac{γ - β }{β} \times 100$

Where:

$∆$ – percentage change in spatial extent;

$β$ – the spatial wetland area for the baseline reference period (2016-2018);

$γ$ – the spatial area for the reporting period.

Known limitations of the data are:

Despite best effort to train the model across the widest range of wetlands possible, there will be types of wetlands and instances of wetland behaviour that will not be adequately captured in a global model. For instance, some ephemeral wetlands are rarely flooded or wet and therefore often missed by satellite datasets. In other cases, the wet part of a wetland may occur under a dense vegetation canopy, which is difficult to assess using Earth Observation data, where the presence of water/moist conditions is not easily detected.

Only regional stratification is applied including strata spanning several countries. Using a finer level of stratification will help improve local/national wetland predictions;
The accuracy of the wetlands map will improve further once cross referenced with more national wetland inventories and ground truthing;
Terrain information from satellite derived DEMs is key input for mapping wetlands globally. The current reference datasets are the 30-meter SRTM DEM which covers the globe from 60oNorth⁰ to 56oSouth⁰, while the region north of 60⁰ north relied on a lower resolution 90-meter DEM model was used. Options for 30-meter DEMs north of 60oN⁰ exists and should be considered in future updates;
Small islands and potentially even entire small island states fall outside the acquisition plan of the Sentinel satellites. As a result, no wetland prediction has been performed for these areas. It will be possible to develop separate models for these missing islands using alternative input satellite data (e.g. using Landsat alone).

Future updates and iterations of the wetlands map will address the above limitations, including a potential shift into a deep learning model to more explicitly reflect temporal and spatial aspects of wetland predictions. Despite limitations with the methodology the production of high-resolution wetland mapping for the entire globe is at the forefront of currently available technology and computing power. It represents a huge step forward towards reporting accurate, statistically robust wetland data.

Mangroves

Description of the method used to measure mangrove area:

Global mangrove area maps were derived in two phases, initially producing a global map showing mangrove area (for 2010) and thereafter producing six additional annual data layers (for 1996, 2007, 2008, 2009, 2015 and 2016) (Bunting et al., 2018). The method uses a combination of radar (ALOS PALSAR) and optical (Landsat-5, -7) satellite data. Approximately 15,000 Landsat scenes and 1,500 ALOS PALSAR (1 x 1 degree) mosaic tiles were used to create optical and radar image composites covering the coastlines along the tropical and sub-tropical coastlines in the Americas, Africa, Asia and Oceania. The classification was confined using a mangrove habitat mask, which defined regions where mangrove ecosystems can be expected to exist. The mangrove habitat definition was generated based on geographical parameters such as latitude, elevation and distance from ocean water. Training for the habitat mask and classification of the 2010 mangrove mask was based on randomly sampling some 38 million points using historical mangrove maps for the year 2000 (Giri et al., 2010; Spalding et al., 2010), water occurrence maps (Pekel et al, 2017), and Digital Elevation Model data (SRTM-30).

The maps for the other six epochs were derived by detection and classification of mangrove losses (defined as a decrease in radar backscatter intensity) and mangrove gains (defined and a backscatter increase) between the 2010 ALOS PALSAR data on one hand, and JERS-1 SAR (1996), ALOS PALSAR (2007, 2008 & 2009) and ALOS-2 PALSAR-2 (2015 & 2016) data on the other. The change pixels for each annual dataset were then added or removed from the 2010 baseline raster mask (buffered to allow detection of mangrove gains also immediately outside of the mask) to produce the yearly extent maps.

Classification accuracy of the 2010 baseline dataset was assessed with approximately 53,800 randomly sampled points across 20 randomly selected regions. The overall accuracy was estimated to 95.25 %, while User’s (commission error) and Producer’s (omission error) accuracies for the mangrove class were estimated at 97.5% and 94.0%, respectively. Classification accuracies of the changes were assessed with over 45,000 points, with an overall accuracy of 75.0 %. The User’s accuracies for the loss, gain and no-change classes respectively were estimated at 66.5%, 73.1% and 83.5%. The corresponding Producer’s accuracies for the three classes were estimated as 87.5%, 73.0% and 69.0%, respectively.

Calculating mangrove changes per country:

Data on mangroves area are available for 1996 and again annually from 2007 to 2020. New annual data will be gradually released. For the purpose of producing national statistics to monitor indicator 6.6.1, the year 2000 has been used as a proxy based on the 1996 annual dataset to align the baseline with that of the surface water dataset.

Annual mangrove extent is compared to this baseline year. Percentage change of spatial extent is calculated using equation 4.

$E q u a t i o n 4 : ∆ = \frac{γ - β }{β} \times 100$

Where:

$∆$ – percentage change in spatial extent;

$β$ – the national spatial extent from the baseline period (2000);

$γ$ – the national spatial extent of any other subsequent annual period.

Note: As river basin delineations does not fully extent into the c

Limitations of the mangrove data:

The mangroves map is a global dataset, and as such, it should not be expected to achieve the same high level of accuracy everywhere as a local scale map derived through ground surveys or the use of very high spatial resolution geospatial data. A global area mapping exercise using consistent data and methods – although supplemented with ground-based data for calibration and validation – for logistical reasons generally requires a trade-off in terms of local scale accuracy. Nonetheless, global maps can be improved locally (or nationally) by adding improved information (in-situ data and aerial or drone data) for training and re-classification.
Several different factors can affect the classification accuracy, including satellite data availability, mangrove species composition and level of degradation.
While the original pixel spacing of the satellite data used for the mapping is 25-30 metres, a minimum mapping unit of approximately 1 hectare is recommended due to the classification uncertainty of a single pixel. The classification errors (in particular omission errors) typically increase in regions of disturbance and fragmentation such as aquaculture ponds, as well as along riverine or coastal reef mangroves that form narrow shoreline fringes of a few pixels.
In general, the mangrove seaward border is more accurately defined than the landward side where distinction between mangrove and certain wetland or terrestrial vegetation species can be unclear.
Striping artefacts due to Landsat-7 scanline error are present in some areas, particularly West African regions due to lack of Landsat-5 data and persistent cloud cover.
Known data gaps in this version (v2.0) of the dataset: Aldabra island group (Seychelles); Andaman and Nicobar Islands (India); Bermuda (U.K.); Chagos Islands; Europa Island (France); Fiji (part east of Antemeridian); Guam and Saipan (U.S.); Kiribati; Maldives; Marshall Islands; Peru (south of latitude S4°), and Wallis and Futuna Islands (France).
As with wetland mapping the production of high-resolution mangrove data for the entire globe is at the forefront of currently available technology and computing power. It represents a huge step forward towards reporting accurate, statistically robust mangrove data which can be updated continuously.

Water Quality: Turbidity and Trophic state

Description of the method used to globally map water quality:

The global dataset to measure water quality for SDG indicator 6.6.1 includes two lake water parameters:

Turbidity (TUR), and
Trophic State Index (TSI).

Both parameters may be used to infer a particular state, or quality, of a freshwater body. Turbidity is a key indicator of water clarity, quantifying the haziness of the water and acting as an indicator of underwater light availability. Trophic State Index refers to the degree at which organic matter accumulates in the water body and is most commonly used in relation to monitor of eutrophication. Turbidity is derived from suspended solids concentration estimates (Binding et al., 2018^[2]) and the Trophic State Index is derived from phytoplankton biomass by proxy of chlorophyll-a (Table 3).

Table 3: Trophic state index and related chlorophyll-a concentration classes (according to Carlson (1977))

Trophic classification	Trophic State Index, Copernicus Global Land Service TSI values	Chlorophyll-a (µg/l) (upper limit)
Oligotrophic	0	0.04
	10	0.12
	20	0.34
	30	0.94
Mesotrophic	40	2.6
Mesotrophic	50	6.4
Eutrophic	60	20
Eutrophic	70	56
Hypereutrophic	80	154
	90	427
	100	1183

The products are mapped at a 300x300 meter pixel resolution capturing monthly data for a total of 4265 lakes and covering two epochs 2006-2010 and 2017-2020. Each lake has individual identification information allowing it to be related to other hydrological datasets. A list of all lake IDs and additional information (location, name – where known, area) is available.

Products in the period 2006-2010 are based on observations from the Envisat MERIS mission, whereas the product 2017-2020 is derived from the OLCI sensors onboard Sentinel 3. Land/water buffer maps as well as ice maps were applied to improve the accuracy of the data. EO-derived water quality parameters are intrinsically difficult to validate, as they strongly depend on the specific lake environment and suitable in-situ data for validation is lacking for most lakes. Still, the general experience of applying EO to derive water quality is that outputs tend to be in accordance with expected spatiotemporal patterns and comparing well to published numbers (Gholizadeh et al., 2016).

Calculating Turbidity and Trophic State Index statistics:

A baseline reference period has been produced comprising monthly averages across 5 years of observations for the period 2006-2010. From these five years of data, 12 monthly averages (one for each month of the year) for both trophic state and turbidity, were derived. A further set of observations are then used to calculate change against the baseline data. These monthly data comprise years 2017, 2018, 2019, 2020 and 2021. The 12 monthly averages for these five years have been derived as used for SDG 6.6.1 reporting.

Monthly deviation of the multiannual baseline is computed using equation 5:

$E q u a t i o n 5 : \frac{M o n t h l y a v e r a g e - M o n t h l y b a s e l i n e}{M o n t h l y b a s e l i n e} \times 100$

For each pixel, and for each month, the number of valid observations has been counted and the number of months where there were monthly deviations, falling in one of the following range of values: 0-25% (low), 25-50% (medium), 50-75% (high), 75-100% (extreme). A corresponding annual deviation synthesis is also produced, and for each target year the number of “affected” lakes relative to the total number of lakes is computed and reported.

A lake is categorized as adversely affected if the combined occurrences of high and extreme changes outweigh those of low and medium changes i.e., (high+extreme) > (low+medium). For turbidity, this rule is applied to the average lake conditions across the year, but for trophic state, which is more event- based, a lake is considered adversely affected if the rule applies for any month within a given year.

The data represent the number of lakes impacted by a degradation of their environmental conditions (i.e. showing a deviation in turbidity and trophic state from the baseline) compared to the total number of lakes within a country. A country or basin’s lake water quality status is labelled as in decline if over 20 per cent of its lakes are affected, based on these criteria. The data is not informing whether a lake is considered to be of good or bad quality, only that a lake water event has occurred and has been recorded. Each event is considered indicative of a degradation in water quality; however, it is important to note that the turbidity and trophic state are included in indicator 6.6.1 as indirect (or proxy) indicators for water quality. These two parameters are not a direct measurement of water quality; however, they perform a very successful proxy role. The proxy parameters are therefore used to alert countries to these events, encouraging countries to investigate why an event occurred and determine if any remedial action is required. You can trace when high and extreme events have event occurred within the advanced analysis of the data.

Known limitations of the water quality data are:

The major limiting factor in satellite-based water quality assessment is the scarcity of available in situ data to support algorithm tuning and validation. Without dedicated field campaigns, automated monitoring stations, and community data sharing arrangements, this is likely to remain a major source of product uncertainty for some years;
Shallow lakes as well as the influence of ice/snow is suspected to add to the observed increase in turbidity levels in the high northern latitudes.

River Flow

Measuring or modelling river flow (discharge):

River and estuary discharge, or the volume of water moving downstream per unit of time, is an essential metric for understanding water quantity within an ecosystem and availability for human use. Countries should provide total annual discharge per major river in order to observe change in river discharge over time.

The river flow sub-indicator measures the changes in the volume of water flowing downstream in rivers and estuaries, also called river discharge. Although the methodology provided for this sub-indicator is flexible, depending on the specificities of countries, the state of their river basins and the national resources available, countries should adhere to the following basic monitoring and reporting guidelines:

Countries are required to provide the total annual discharge for all major rivers and monitor changes in river discharge across years.
Discharge data from each major river monitored should be collected at least once per month. This data should then be averaged to obtain an annual average discharge per river.
Each basin should have a minimum of one sampling location, at the point where its water exits into another basin or at the exit point from major tributaries.

The in-situ monitoring methods for river discharge are flexible and can include gauging stations, current meters, or even modelled discharges from hydrological/hydraulic models (preferably complemented with in-situ data, where possible, to ensure accuracy).

While the river flow sub-indicator is primarily intended as being measured in-situ, with techniques including gauging stations and discharge meters. The availability of in-situ observations is spatially heterogeneous and scarce in large parts of the world. Furthermore, large scale monitoring networks are expensive and, in many cases, impractical, particularly for large scale or subsurface processes such as groundwater dynamics. River flow (and ground water) cannot be directly observed from space, but they can be simulated by combining Earth Observations and numerical simulations.

In response to this situation a modelling approach has been adopted for the global reporting on river flow changes. The approach is based the DHI Global Hydrological Model (DHI-GHM) which provides historical, real-time, 10-day and seasonal forecast simulations at the global scale. DHI-GHM is comprised of a distributed gridded rainfall-runoff model with a spatial resolution of 0.1° and an agile kinematic routing model that moves water between model grid cells and sub-catchments and in the river system. Model output includes gridded hydrological variables, such as soil moisture content and different runoff components, and discharge at more than 1 million river points globally (Murray et al., 2023).

From the DHI-GHM model, monthly river discharge data since year 2000 are obtained for all relevant level 12 hydro basins and used to estimate annual data of minimum and maximum river flow. River flows from the outlets of the level 12 basins that serve as outlet basins of the higher levels are aggregated to calculate the flows of higher-level basins. The total annual min./max. discharge for countries are computed as the sum of the discharge at all river outlets falling within a given country.

To calculate percentage change in river, flow a 20-year baseline period for annual minimum and maximum flow is calculated nationally and for each sub-basin level. This baseline period is used to calculate the percentage change of discharge for any subsequent 5-year period (cf. equation 6).

$E q u a t i o n 6 : ∆ = \frac{γ - β }{β} \times 100$

Where

∆ = Percentage Change in River Discharge

β = historical 20-year reference discharge i.e. the median of the annual minimum and maximum discharge during the 2000-2019 period.

γ = the median of the annual min./max. discharge over the 5-year reporting period (e.g., 2017-2021).

Changes in river flow are calculated nationally and sub-nationally for all hydro basin levels.

Important considerations:

The sections below describe key considerations for monitoring discharge and provides criteria for discharge to complement the global data currently generated for Indicator 6.6.1.

Common in-situ monitoring methods: There are a variety of methods for monitoring discharge in situ and selection should be based on the size and type of the waterbody, terrain and velocity of water flow, the desired accuracy of measurement, as well as finances available. Two the most common and accessible approaches are gauging stations and using current meters. In many countries, gauging stations are the most prevalent means for measuring river discharge as they allow even for continuous and often real-time monitoring. These are fixed locations along a river or estuary where the change in water surface level (stage) is monitored at locations where a unique relationship exists between stage and flow and a so-called rating curve can be produced. Water surface height (stage) is captured frequently, and the discharge estimated, most often at monthly intervals but in many places, this is available at daily intervals or even continuously. Current meters and other instruments can be used to monitor flow and calculate discharge. For example, propeller, pygmy or electromagnetic current meters are often used to measure velocity and can be used in conjunction with cross-sectional area methods to obtain flow rates. Acoustic Doppler Current Profiler’s (ADCPs) are widely used for larger rivers/estuaries to accurately measure bed depth, velocity, and discharge. They are often attached to boats and dragged along a waterbody, but permanent installations can also be found, sending out acoustic waves and measuring acoustic reflectance. Meters and instruments like ADCPs are significantly more costly than other methods of measurement and require skilled operators and good maintenance programmes. However, in larger rivers they may be the most appropriate option, especially during high flow conditions.
Location of Monitoring: The chosen monitoring method may dictate where along a river or estuary the discharge is captured. For example, if fixed weirs are in place, monitoring will always take place here. Since in situ discharge monitoring can be time and cost-intensive, choosing strategic locations which represent a whole river or estuary is recommended. The minimum monitoring effort is to locate one flow measuring site within proximity to each basin’s exit (into another basin). In addition, monitoring at the exit point from all major tributaries adds a substantial level of information. Where there is a local impact on discharge due to human influence, then it is recommended to monitor flow upstream and downstream of these areas so that the overall situation can be managed.
Frequency of Monitoring: The quantity of water in a river or estuary can change rapidly in response to rainfall and weather patterns. The more data on discharge there is, the higher the accuracy is of that discharge data. However, again it is important to focus efforts and choose a strategic frequency for monitoring. Data on discharge should ideally be collected at a given location once a month at minimum (ideally at a daily frequency) and this data can then be used to determine annual and long-term trends. The quantity of water in estuaries may be significantly influenced by tidal inflows, thus this indicator is limited to the freshwater inflows to the estuary from the upstream river.
Modelling Discharge: In addition to in situ monitoring which always is impacted by all forms of flow moderation, storage or abstractions upstream, discharge may also be modelled from one of the many available models which use climatic and land-use data, amongst other data, to estimate both natural and present-day flows. Globally hydrological model applications are available and in some countries these or similar models have been developed for the local context and are calibrated using real measured data. It is recommended that modelled discharge data is complimented by measured in situ data wherever possible to ensure accuracy. Conceptual hydrological models for flow and discharge estimation are normally less amenable to detecting the flow impacts of minor land-cover changes over time as the models are calibrated on historical flow data and associated land-use conditions.

Groundwater

Measuring quantity of groundwater within aquifers:

The changes to the quantity of groundwater within aquifers is important information for many countries that rely heavily on groundwater availability. For the purposes of SDG indicator 6.6.1 monitoring the changes to groundwater levels gives a good indication of changes to the water stored in an aquifer. Furthermore, only significant ground water aquifers, that can be seen as individual freshwater ecosystems will be included in the reporting.

Important considerations:

The sections below describe some key considerations for monitoring groundwater changes for SDG indicator 6.6.1:

Location of Monitoring: Measuring the level of groundwater within an aquifer is done through the use of boreholes. One of the challenges in setting up monitoring is choosing the location of boreholes which will adequately represent the total groundwater situation for an aquifer. The number of boreholes that need to be monitored cannot be prescribed because the distribution of groundwater can be variable depending on the location and characteristics of aquifers. It is recommended that sufficient boreholes to characterise the area should be monitored, with the capacity of the country being a factor in deciding how many would best represent the area. It is highly recommended that data should be taken from observation boreholes / monitoring boreholes (these are boreholes which are not equipped with pumps). Data from used (pumped) boreholes should be avoided. In case a pumped borehole needs to be used for measurements, then it is crucial to allow for a sufficiently long recovery period in which the borehole is not used so that the groundwater level in the borehole can stabilise prior to any measurement.
Frequency of Monitoring: Groundwater levels change as a result of changes in groundwater recharge (affected by climate conditions, and land use) and by anthropogenic removals from the system (groundwater abstraction). Seasonal and wet/dry cycle influences need to be understood and hence monthly monitoring is optimal, but collection at least twice per year, in the wet and dry seasons, is necessary.
Criteria for Indicator 6.6.1 Data: Groundwater quantity data provided to the custodian agency(s) will be quality checked to ensure data integrity. Collection of groundwater level data generates statistics that are a proxy to the quantity of groundwater in an aquifer over time. In order to examine this change over time, percentage change in groundwater level will be generated and validated between the custodian agency(s) and the country. Calculating percentage change at a national level requires the establishment of a common reference period for all aquifers, which can either be based on historical groundwater level data (preferred) or modelled data if available. In cases where these are unavailable, a more recent period can be adopted to represent the ‘baseline’ or reference period. Countries should provide the annual level of groundwater in order to observe change in aquifer volume over time. A data collection table is provided in the monitoring methodology as an annex.

National Indicator Score

The overall national summary score, which assesses progress toward achieving SDG Target 6.6 on protecting and restoring freshwater ecosystems, is derived by integrating the individual sub-indicators. The score is based on the evaluation of changes in the sub-indicators and based on the One Out, All Out (1OAO) principle that is if one of the sub-indicators is in decline the national summary score would be considered in decline, and in strong decline if there are two or more declining sub-indicators. If no sub-indicators are declining the following applies: stable, when all sub-indicators are neither declining nor improving; improving, when one or two sub-indicators is improving and greatly improving when there are more than two improving sub-indicators.

¹

European Space Agency, Sinergise (2021). Copernicus Global Digital Elevation Model. Distributed by OpenTopography. https://doi.org/10.5069/G9028PQB. Accessed: 2024-03-24 ↑

²

Binding, C., Stumpf, R.P., Schaeffer, B.A., Tyler, A. and Hunter, P., 2018. Chapter 2: Introduction to Deriving Water Quality Measures from Satellites. Reports and Monographs of the International Ocean Colour Coordinating Group (IOCCG), 17, pp.15-28. ↑

Hesin internet-kagin er ikki longur stuðlaður

Rætta uppseting: 6.6.1