© OpenStreetMap contributors
- Add External Layer
- Upload Shapefile
- Layer Tutorial
- Close
Use the checkbox () to show or hide a layer.
Use the radio buttons () to select a layer to use with the identify function.
- Layers
- Opacity
Available layers
(ctrl+c)
Search for marine data across UK organisations
- API
- How-To
- About
- Contact MEDIN
- Share
Metadata: 2007 Scottish Natural Heritage (SNH) Luce Bay (Scotland) sublittoral broad-scale mapping survey
Abstract:
Luce Bay and Sands SAC is a broad, shallow embayment on the south-west coast of Scotland. The site has been designated for four qualifying Annex I habitat types Large shallow inlets and bays, Sandbanks which are slightly covered by seawater at all times, Intertidal mudflats and sandflats and Reefs. The aim of this project was to derive broadscale maps for the intertidal and subtidal Annex 1 habitats within the SAC. Because of the size of Luce Bay, it was decided that remote sensing supported by ground truthing of the data would be the most cost-effective way of mapping the habitats within the site. The intertidal would be surveyed using satellite imaging and the sublittoral environment would be surveyed using acoustic survey. Ground truthing would be undertaken by direct observation, where access allowed, in the intertidal and by a combination of video observation and grab sampling in the sublittoral.
Data holder:
NatureScot (HQ Inverness)
Click on the red button for resource contact details:
- Click on the red button for resource contact details
| Other details | ||
| Internal code | Internally assigned metadata identifier | 11335 |
| Title | The title is used to provide a brief and precise description of the dataset such as 'Date', 'Originating organisation/programme', 'Location' and 'Type of survey'. All acronyms and abbreviations should be reproduced in full. | 2007 Scottish Natural Heritage (SNH) Luce Bay (Scotland) sublittoral broad-scale mapping survey |
| File Identifier | The File Identifier is a code, preferably a GUID, that is globally unique and remains with the same metadata record even if the record is edited or transferred between portals or tools. | 3f7776fb409ec2edb3c1a7eec8cb6a4f |
| Resource Identifier | This is the code assigned by the data owner. | GB-SCT-SNH-ME-000110-MRSNH01000000004-LUCE |
| Resource type | The resource type will likely be a dataset but could also be a series (collection of datasets with a common specification) or a service. | dataset |
| Start date | This describes the date the resource starts. This may only be the year if month and day are not known | 2007-05-19 |
| End date | This describes the date the resource ends. This may only be the year if month and day are not known | 2007-10-09 |
| Spatial resolution | This describes the spatial resolution of the dataset or the spatial limitations of the service. | inapplicable |
| Frequency of updates | This describes the frequency with which the resource is modified or updated i.e. a monitoring programme that samples once per year has a frequency that is described as 'annually'. | notPlanned |
| Abstract | The abstract provides a clear and brief statement of the content of the resource. | Luce Bay and Sands SAC is a broad, shallow embayment on the south-west coast of Scotland. The site has been designated for four qualifying Annex I habitat types Large shallow inlets and bays, Sandbanks which are slightly covered by seawater at all times, Intertidal mudflats and sandflats and Reefs. The aim of this project was to derive broadscale maps for the intertidal and subtidal Annex 1 habitats within the SAC. Because of the size of Luce Bay, it was decided that remote sensing supported by ground truthing of the data would be the most cost-effective way of mapping the habitats within the site. The intertidal would be surveyed using satellite imaging and the sublittoral environment would be surveyed using acoustic survey. Ground truthing would be undertaken by direct observation, where access allowed, in the intertidal and by a combination of video observation and grab sampling in the sublittoral. |
| Lineage | Lineage includes the background information, history of the sources of data, data quality statements and methods. | For intertidal mapping, remote sensing offered a specific advantage in its ability to obtain temporal and spatial information at scales unmatched by other survey methods such as aerial photography. Mapping of marine intertidal zones required remotely sensed data at high spatial, spectral and radiometric resolution to match the scale of the variations in habitat types and to offer the ability to discriminate different habitat types. Two complementary acoustic techniques were used to survey the subtidal element of Luce Bay SAC; a swathe bathymetric system and an Acoustic Ground Discrimination System (AGDS). The swath system was used as this not only gave bathymetry across a swath of 10 to 12 times water depth (important in the shallow waters of Luce Bay) but also side scan-quality images. The system had accurate motion sensing so all images were of high quality and corrected for motion and tide. The system used was the GeoSwath Plus from GeoAcoustics. The second acoustic system, RoxAnn AGDS, was run concurrently and the information used to assist the interpretation of the swath data. The biotope records from the separate biological surveys were used as ground truth data upon which to base the analysis and interpretation of the remote sensing and acoustic images. [for details relating to the biological ground truthing for the sublittoral see the final paragraph in this description] The purpose of the analysis and interpretation was to provide distribution maps of biotopes and seabed features within Luce Bay to aid future management of the site. Survey logistics: Satellite image acquisition took place at a suitable low water spring tide during summer 2007 to maximise the area of coverage. Field measurements to support the geometric and atmospheric pre-processing of the satellite imagery were collected from the 26th to the 30th May 2007. The acoustic survey was undertaken from the 19th May to the 1st June 2007. Subtidal groundtruthing was undertaken from the 9th to the 12th July 2007, with two further gap filling days on the 6th August and 9th October 2007. The intertidal biological ground truthing survey took place from the 12th to the 21st June 2007. Satellite imaging Satellite imaging - Establishment of ground control points: The QuickBird imagery was supplied by DigitalGlobe Inc. in geocorrected form to a 1:50,000 scale accuracy. To check the accuracy of the imagery, 25 prominent and permanently located sites, which were discernable on the imagery, were measured for their precise location using a EGNOS-enabled 12 channel Magellan Marine GPS unit. GPS station positions were recorded for between 3 to 5 minutes, which, with EGNOS availability, gave a positional accuracy of approximately 2 m, within one pixel of the true position of the multispectral imagery. Positions were recorded in UTM WGS84 coordinates. The same GPS unit was used to accurately locate all other land and water-based measuring sites. Satellite imaging - Measurements of land field targets for atmospheric correction: Atmospheric correction of the imagery was undertaken using the empirical line method which provides an accurate method for atmospherically correcting high spatial resolution satellite imagery (Karpouzli and Malthus, 2002). Thirteen large and relatively homogeneous land targets of varying brightness were measured for their spectral reflectance properties using a GER 1500 spectroradiometer. The targets measured ranged from tarmac, concrete, grass and sand. The methods employed are outlined in detail in Bates et al. (2004a). Care was taken to ensure that targets of varying brightness were included in each separate sub-image covering the Luce Bay area as presented above, in case acquisition of the three separate images occurred on different days and under different atmospheric conditions. Satellite imaging - Satellite image processing: several images were used to represent near complete coverage of the Luce Bay intertidal and shallow subtidal regions. Satellite imaging - Geocorrection: All satellite images were acquired in WGS84 UTM Zone 29 projection. All subsequent processing was undertaken in this projection, with data products re-projected at the end of the analysis to the GB OS grid. Qualitative comparison of field-measured GPS data points with the original images supplied by DigitalGlobe indicated that the original data supplied were accurate only to within 4 to 20 m of 'true' location, dependent on the specific image studied. To ensure accurate geopositioning and the correct matching up of the images with each other, a geometric correction was performed on the large sun-glinted image (based on the field measured GPS points) and all other data sets were co-registered to it, using up to 30 registration points for each image. The results led to the reduction of the positional error to less than the dimension of a single pixel (Total RMSE = 2.3 m). A simple linear shift was performed with nearest neighbour resampling. Satellite imaging - Atmospheric correction: To deal with sun glinted imagery, relationships were developed between the calculated ground-based QuickBird reflectance values and related values extracted from the corresponding pixel locations in the multispectral QuickBird image datasets. The relationships for all four bands were all highly linear (R2 values > 0.97, example for the image shown in Table 3.8). These empirically derived relationships were used to atmospherically correct QuickBird imagery to percent reflectance (Karpouzli and Malthus, 2002). Subsequent images used in the biotope classification were corrected to reflectance using pseudo-invariant targets of varying brightness visible in the corrected imagery and in each image in a similar approach to those using the ground targets outlined above. Satellite imaging - Masking: To remove land areas, and in the absence of high resolution Lidar DEM data, the corrected QuickBird image dataset was masked to a Scottish coastline vector dataset. The vector data was manually adjusted to the visible highest astronomical tide, where possible, including the splash zone/strandline as ground truthing field survey points were included in this region. Areas of cloud and cloud shadow were also removed from the imagery by manually digitising around the affected areas. Satellite imaging - Biotope classification: The optical image datasets were classified to marine biotopes using supervised maximum likelihood classification. Separate training areas were delineated on each image for each biotope. Combined classifications were produced for the intertidal and subtidal biotope components. The ground truthing survey data were used to provide the basis for training the classification process. All ground truthing station data were first assigned a biotope class and a separate ID code. Using this information, training areas were then delineated and assigned on each image component for each biotope except for those biotopes for which there were only 1~2 stations identified. Biotopes that were spectrally indistinguishable were combined in common biotope class definitions on the basis of their hierarchical biotope groupings (Connor et al., 2004). A 7x7 pixel majority filter was also used to remove isolated mis-classified pixels. Acoustic survey Acoustic survey - Track spacing: Acoustic tracking was carried out over the whole of the survey area at 1 km spacing and three priority areas were surveyed using closer track spacing to collect data to ensure at least 100% coverage. Survey lines were planned in a GIS system, and transferred to the vessel electronic chart plotter, which allowed the vessel position to be displayed along the survey lines. The vessel skipper could then follow the course in real time and follow the survey lines as closely as possible. Acoustic survey - Swath bathymetry/side scan sonar: The swath system used by Envision was GeoAcousticsâ?? GeoSwath Plus. It is an interferometric system operating at 250 kHz that uses the phase content of the signal to measure the angle of the wave front of the returning echo. The depth of the seabed reflecting surface is calculated from this angle and elapsed time of the signal return. The system also measures signal amplitude and sidescan style images are derived from this information. The system offers a good resolution from which accurate and detailed bathymetric models can be produced. Acoustic survey - RoxAnn AGDS System: Envision used a RoxAnnâ?¢ GroundMaster AGDS operating at 50 kHz. Acoustic ground discrimination systems (AGDS) are based on single beam echo sounders and, apart from determining depth, are designed to detect different substrata by their acoustic reflectance properties: hard surfaces produce strong echoes whilst soft surfaces result in a weak signal. Additionally, rough surfaces will produce an echo that decays slowly (a property termed â??backscatterâ??) whilst flat surfaces result in a rapid decay of the signal. These properties can be used to discriminate broad categories of sea floor habitats (Foster-Smith et al., 2000; Foster-Smith and Sotheran, 2003). The acoustic data, together with GPS data, are logged onto a laptop and the systems are portable and self-contained. The data was analysed using image processing and GIS (Sotheran et al., 1997) after carrying out detailed data quality assurance procedures. Acoustic survey - Performance of the acoustic systems: The GeoSwath Plus operated at 250 kHz and power setting two and ping length between 50 and 70 m depending on water depth and line spacing; this gives a maximum swath width between 100 and 140 m. The AGDS operated at 50 kHz. Basic processing (during which data are corrected for GPS offsets, tide and variations in speed of sound, gridded and mosaicked) cannot be undertaken with the GeoSwath system (on previously collected data) at the same time as data are being acquired. The raw data were backed up every evening and processed using a second land based PC, enabling estimates of coverage and data quality to be made on a day to day basis and allowing for daily survey plans to be adapted. The AGDS data were processed on a day-to-day basis. The data were consistent between days and it was judged that there was no requirement for standardisation between days. The data were imported into MapInfo for display as a thematic map of roughness and hardness. Acoustic survey - RoxAnn AGDS: The RoxAnn data were edited on a day-to-day basis. The data were cleaned (removal of depth spikes, records when the vessel was stationary and where there were zero values for depth). The data were imported into MapInfo and displayed showing incremental values in E1 (roughness) and E2 (hardness) and records removed where there was any doubt about values (as compared to surrounding tracks). The data were acquired in latitude and longitude, WGS84 datum. The data were then converted to have positions in OSGB36 datum. The daily datasets were compared using scattergrams and there appeared to be no obvious shifts in the patterns or overall values between days. Acoustic survey - Swath bathymetry: survey lines were acquired with the swath system. Acoustic survey - Characterisation of the acoustic data: The swath acoustic data were prepared to show topographic and sediment characteristics. The digital bathymetric data were gridded in Surfer (Sotheran et al., 1997). A relief map was then created which was colour-coded to depth and â??sun-illuminatedâ?? to show topography. This revealed larger topographic features (small and large sand waves, irregular scour marks and rugged rocky terrain). The sidescan images were mosaiced. However, most of the tracks were widely spaced and mosaicing was used primarily to georeference the images. Sidescan images showed features defined primarily by changes in acoustic hardness (soft, hard ground; rugged and acoustically variable ground; ribbons of alternating hard and soft ground). Both relief maps and sidescan images were imported into MapInfo GIS. The AGDS were coloured according to hardness and softness and superimposed over the bathymetric and sidescan images. Lastly, the groundtruth sample data were superimposed over all the other layers. The tracks were then inspected and sections of track were characterised by the habitat features. Additionally, the orientation of any feature that showed marked directionality (waves, ribbons) was marked by a line crossing the track. The interpretation is dependent upon the judgement of the analyst (as is usual for the interpretation of swath images) and is, therefore, subjective. Nevertheless, the correspondence between interpreted seabed types and the biotopes (or grouped biotopes) can be assessed to see if they match what could reasonably be expected. The correspondence between the biotope groups and the features indicates that many of the broad, sediment-based biotopes can be associated with a range of features. In particular, the mixed sediment biotopes were spread between cobbley gravel and shelly muddy sand sediments. The muddy sands were also very variable with an element of coarse material (shelly gravel) being indicated in many samples. All the biotopes that require a hard substratum were, however, found on cobbley gravel and bedrock. Thus, the correspondence might be expected allowing for some cross-over between biotope and seabed type. Acoustic survey - Classification of the Acoustic Ground Discrimination System (AGDS) in terms of biotopes: The tracks were mostly too widely spaced for interpolation to be performed on the AGDS point track data. Adjacent tracks often showed quite a marked difference in values indicating a marked change in seabed between the two tracks. Mathematical interpolation performs poorly in these circumstances. Instead, a supervised classification can be performed on the track data leaving the areas between tracks blank. The point AGDS data were transformed into a raster image for hardness, roughness and depth which were imported into Idrisi(tm) where standard supervised classification routines were applied. The biotope and grouped biotope sample data were used for ground truthing. It is important to understand that the classification gives a prediction of the distribution of biotopes (or groups) and is prone to classification errors. Nevertheless, the agreement between the ground truth data and the classified AGDS data was fair for the grouped biotopes (69% correspondence and a Kappa measure of 0.60). The full list of biotopes resulted in a lower correspondence (65%, Kappa 0.55), but this is only to be expected given that many of the biotopes were represented by single records. More general biotopes (i.e. the grouped biotopes) are much more robust for classification. Acoustic survey - Combined interpretation of track characteristics and AGDS classification: The final interpretation required two steps: firstly, the evidence from both the track characteristics and the AGDS classification was assessed to determine boundaries between biotopes and, secondly, areas of similar seabed were encompassed by digitised polygons that were drawn to produce a continuous coverage of the survey area. Both steps depend upon the judgement of the analyst and are not strictly objective processes. It should be noted that the assigned biotopes (or biotope groups) sometimes appear to conflict with the polygon. This occurred when there was conflict between the evidence from the sidescan/RoxAnn and the groundtruth samples. It was often found that the full description contained in the records indicated that the habitat was mixed. This agreed with the hard return for the sidescan and AGDS and it was decided that the evidence from the acoustic data should take precedence over the point sample data in these circumstances. Biological ground truthing survey Biological ground truthing survey - Intertidal survey: The intention for the intertidal ground truthing survey was to identify a number of locations from the satellite image for biotope identification that would then be used in the classification of the remote image. However, the timing of the satellite image acquisition was such that the images were not available before the intertidal survey period. To ensure a wide range of biotopes were sampled across the whole of Luce Bay, intertidal samples were selected from observation of the existing aerial images of the site and also from examination of Ordnance Survey maps. In total 176 locations were sampled. At each target location information was collected on the position from GPS, the main physical characteristics and dominant epibiota for rocky shores, and, for sediment shores, eight pooled 10.1 cm diameter corer samples were taken for the identification and enumeration of infauna. Digital photographs were taken of the main habitats and species found. Biological ground truthing survey - Subtidal survey: The drop-down video and grab surveys provided information on the species and communities present in the subtidal areas of interest, and enabled identification of biotopes for subsequent mapping and assessment of the principle conservation interests for the features across the site. A total of 76 video and 13 grab stations were sampled. For reef areas, data on the species and communities present were collected using a dropdown video camera and photography system. This method was ideal for identifying conspicuous epifauna and flora on rocky habitat. The methods followed those outlined in Procedural Guidelines 3 to 14 in situ Survey of Sublittoral Epibiota using Towed Sledge Video and Still Photography and 3 to 5 Identifying Biotopes using Video Recording, from the Marine Monitoring Handbook (Davies et al., 2001). Note that Procedural Guideline 3 to 5 was updated in Sotheran and Foster-Smith (2004). The video was collected on a Bowtech BP-L3C-2 digital video camera and recorded on a SONY mini-DV recorder and then transferred to computer hard drive for subsequent interpretation. Grab sampling was undertaken from a small Van Veen Grab with replicate samples taken to equate to 0.1 m2 at each ground truth sampling location (4 samples pooled for biology and one for Particle Size Analysis). The advantage of a combined video and grab system was that both epifauna and infauna were recorded from the same location, making the identification of sediment communities more reliable than either method individually. For the subtidal survey camera work, drops of up to 5 to 10 minutes bottom time were used to collect information on the main substrata and to identify conspicuous epifauna and assess abundance across the SAC. The exact positioning of each drop depended on the heterogeneity of the seabed as gauged from the geomorphological work outlined above, but aimed to provide adequate coverage across the whole of the SAC. |
| Additional information | This describes relevant references to the data e.g. reports, articles, websites plus other useful information not captured elsewhere. | ERT (Scotland) Ltd (2011) Broadscale mapping of marine features within the Luce Bay and Sands Special Area of Conservaion. Scottish Natural Heritage Commissioned Report No.471. |
| Related keywords | ||
| Keyword | General subject area(s) associated with the resource, uses multiple controlled vocabularies | Marine Environmental Data and Information Network |
| General subject area(s) associated with the resource, uses multiple controlled vocabularies | NDGO0005 | |
| Keyword title | data.gov.uk | |
| Keyword | General subject area(s) associated with the resource, uses multiple controlled vocabularies | Geology |
| General subject area(s) associated with the resource, uses multiple controlled vocabularies | Habitats and biotopes | |
| General subject area(s) associated with the resource, uses multiple controlled vocabularies | Species distribution | |
| General subject area(s) associated with the resource, uses multiple controlled vocabularies | Habitat characterisation | |
| General subject area(s) associated with the resource, uses multiple controlled vocabularies | Habitat extent | |
| General subject area(s) associated with the resource, uses multiple controlled vocabularies | Zoobenthos taxonomic abundance | |
| General subject area(s) associated with the resource, uses multiple controlled vocabularies | Zoobenthos taxonomy-related counts | |
| Geographical coverage | ||
| North | The northern-most limit of the data resource in decimal degrees | 54.857219650858 |
| East | The eastern-most limit of the data resource in decimal degrees | -4.3574523925781 |
| South | The southern-most limit of the data resource in decimal degrees | 54.624145878862 |
| West | The western-most limit of the data resource in decimal degrees | -5.0166320800781 |
| Regional sea | Irish Sea | |
| Responsible organisations | ||
| Role | The point of contact is person or organisation with responsibility for the creation and maintenance of the metadata for the resource. | pointOfContact |
| Organisation name | NatureScot (HQ Inverness) | |
| data_supply@nature.scot | ||
| Role | The owner is the person or organisation that owns the resource. | owner |
| Organisation name | Scottish Natural Heritage (SNH), Headquaters | |
| Individual name | Scottish Natural Heritage (SNH) | |
| Position name | Data Manager | |
| Phone | 01463 725000 | |
| data_supply@nature.scot | ||
| Role | The custodian is the person or organisation that accepts responsibility for the resource and ensures appropriate care and maintenance. If a dataset has been lodged with a Data Archive Centre for maintenance then this organisation is be entered here. | custodian |
| Organisation name | NatureScot (HQ Inverness) | |
| Position name | NatureScot Data Manager | |
| data_supply@nature.scot | ||
| Role | The originator is the person or organisation who created, collected or produced the resource. | originator |
| Organisation name | ERT (Scotland) Ltd | |
| Position name | Marine Environmental Consultancy | |
| dassh.enquiries@mba.ac.uk | ||
| Role | The distributor is the person or organisation that distributes the resource. | distributor |
| Organisation name | Data Archive for Seabed Species and Habitats (DASSH) | |
| Individual name | Esther Hughes | |
| Position name | Data Manager | |
| Phone | 01752 633102 | |
| dassh.enquiries@mba.ac.uk | ||
| Dataset constraints | ||
| 20.1 Limitations on Public Access - Access constraints | This states `otherRestrictions` from ISO vocabulary RestrictionCode and is an INSPIRE/GEMINI requirement. | otherRestrictions |
| 20.2 Limitations on Public Access - Other constraints | no restrictions to public access | |
| 21.1 Conditions for Access and Use - Use constraints | This states `otherRestrictions` from ISO vocabulary RestrictionCode and is an INSPIRE/GEMINI requirement. | otherRestrictions |
| 21.2 Conditions for Access and Use - Other constraints | This states any constraints on use of the data. Multiple conditions can be recorded for different parts of the data resource. If no conditions apply, then `No condtions apply` is recorded. This uses free text. | Not for navigational use; SNH copyright data which is available for re-use under government licence terms: http://www.nationalarchives.gov.uk/doc/open-government-licence/version/3/ |
| Available data formats | ||
| Data format | Format in which digital data can be provided for transfer | Database (Marine Recorder) |
| Format in which digital data can be provided for transfer | Documents (Published report) | |
| Format in which digital data can be provided for transfer | Geographic Information System (shapefiles) | |
| Version info | ||
| Date of publication | The publication date of the resource or if previously unpublished the date that the resource was made publicly available via the MEDIN network. | 2011-01-01 |
| Harvest date | The date which this record has been (re)harvested from the provider. | 2026-04-19 |
| Metadata date | The date when the content of this metadata record was last updated. | 2025-06-04 |
| Metadata standard name | The name of the metadata standard used to create this metadata | MEDIN |
| Metadata standard version | The version of the MEDIN Discovery Metadata Standard used to create the metadata record | 3.1.2 |