User Guide - Scottish Natural Heritage
Contents
1. 2 NCHS Main View 56 59 Scale 1 3 10 1 79 000 BA DS D cor na NIS Cra Ca al HCh S squares e Coastline Stratification LL BE New themes can be added to the NCMS Main View in the usual way Standard ArcView resizing tools can be used to enlarge or zoom in on the View If a project is saved with the NCMS application active the NCMS extension will automatically be loaded when the project is next opened National Countryside Monitoring Scheme Visualisation and Analysis System To unload the application uncheck the tick next to NCMS v1 5 in the list of extensions and click OK This will cause the NCMS Main View to disappear so ensure that your project is saved first Before using the application it is recommended that you choose the printer you intend to use by selecting Print Setup from the File menu This will help ensure that margins are correctly set in ArcView layouts For information on any remaining known issues with the software consult the ncmsissues pdf file on the CD ROM National Countryside Monitoring Scheme Visualisation and Analysis System 5 Tutorial This chapter introduces the main features of the application in the form of a step by step guide A reference for each of the NCMS menu items can be found in Chapter 6 The NCMS application has two main functions e to display and analyse NCMS land cover data for a selected NCMS sample square and2. OF Cancel Options are available under four headings Stock The estimated amount of each feature in the Analysis Region will be calculated for each date selected One two or three dates may be selected Net Change The estimated change in area or length of each feature over the selected time period will be calculated Only one period can be selected so if estimates for 40s 70s and 70s 80s say are required two separate analyses will have to be run Features Selects whether results for areal and or linear features are required Interchange The estimated interchange between different features during the selected time period will be calculated e g how much grassland changed to arable Interchange can be estimated only for areal features and only one time period can be selected for each analysis run Note that the calculation of interchanges may take several minutes on slower computers Finally select whether standard errors or 95 confidence intervals are required The Define Groups button activates the Define Groups dialog described in Section 6 3 Clicking on OK will display the Analysis Options dialog 6 4 2 Analysis options The Analysis Options dialog Figure 6 8 is split into two sections Analysis Method offers four choices determining how results should be calculated These determine which squares are used to calculate the estimates and associated standard errors and hence confidence intervals See Ap
3. and select a square from the list or e click on a square in the NCMS Main View If necessary enlarge the View window and zoom in on part of the map to make it easier to select the square you want Once selected the square should be highlighted in the View Only one square can be selected at a time Leave the rest of the View Square dialog unchanged and click on OK Once processing is complete an ArcView layout and a report window should be displayed The report window will be on top These should be similar to those shown in Figure 5 2 and 5 3 Figure 5 2 The interchange report for the square selected in Figure 5 1 ib iji h E In ET harp Regul igli rl jas ee ees 532 53 Zi49gra 19408 te 19505 Interchanze nr chang 11 37 cacas Er suicida co FICS nagh gian Fai ral USA E bluzkst rera Eqathar baatiar mio ER 32 haathar reo cand atu parlar aT haaiher uu clan tlarisiqmhse Lheaibhe 125 313 hilari ru arid ati lari el Pee ib akku LEN bluzkst rera Equthar zr grazilud 16 13 blazkat TETS grace Haricrt mos hagihar TM rara geram ani dl blank mie herlhe EE quera sovs tlaricei mos hasha 311 utra soa xaT IM Creal ee Fai mr laal Lal bluzkst rera bamar banh 192 haathar reo crdland elder mos grazz 1 48 IENE lab mol 13H hallar iE clan eee Bre 133 bluzkst rer bamar ars grund 1 La haalher mo celand IFEN 1 14 mapal iaria a cug phat Chd 1 08 hlinik mar Hahi A Mth ad 108 TE cover larieri mEes paaa
4. You may include several polygons either by clicking in each polygon while holding down the shift key or by drawing a boundary around the polygons to be included Once the required selection s have been made click on the OK button A dialog will appear prompting you to save the Analysis Region theme Specify a file name and a location for the theme which will then be added to the NCMS Main View b By drawing area s with mouse Selecting this option will display a dialog asking where the new Analysis Region theme should be saved Specify a file name and a location for the theme The drawing tools provided can be used to draw the boundaries of the new Analysis Region Figure 6 5 Once the required area or areas have been drawn click the OK button to save the Analysis Region 21 National Countryside Monitoring Scheme Visualisation and Analysis System Figure 6 5 Draw Analysis Region dialog 2 Draw Analysis Region Draw the boundary of the areale using the tools below Click OK when Finished c From current selection of active theme Choose From current selection of active theme if you have already selected the required region in the NCMS Main View You will be prompted for a file name and location to save the Analysis Region theme 6 3 Define Feature Groups Related land cover features such as the various types of woodland can be grouped together for analysis Although estimates for a group of features c
5. e to enable estimates of land cover and land cover change to be calculated for a user defined geographical area referred to as the Analysis Region These functions are accessed though the NCMS menu and by interacting with the NCMS Main View 5 1 Displaying land cover in a sample square To display land cover within an NCMS sample square first select View Square from the NCMS menu if the NCMS menu is not visible make sure the NCMS Main View window is active You should see the dialog shown in Figure 5 1 Figure 5 1 The NCMS Square Viewer Dialog 2 NCHS Square Viewer Choose square fram list or select square fram NOMS View using the Select Feature tool Selected MCMS Square 249qra Land Cover Chang e 15940 EE Aeris sens v 1980s 405 805 Displa Features M Areal wv Linear Format M Layout Calculate Areas E Files ta be created A4 Igra View 43gra Layout Layout Template Ad OF Cancel Each square has a unique code a number followed by either two or three letters The letters refer to the NCMS region or district containing the square The square selected in Figure 5 1 is in Grampian A list of the districts and their codes can be found in Appendix 5 National Countryside Monitoring Scheme Visualisation and Analysis System You can select the square to be viewed in two ways Either e click on the drop down box at the top of the dialog next to Selected NCMS Square
6. new Analysis Region i e using a pre defined set of options The analysis file contains e the name and location of the Analysis Region shapefile the definition of any areal and linear feature groups if relevant output options and specific strata settings if these were defined in Advanced Analysis Options Retrieve Analysis on the NCMS menu is used to load an existing analysis file You will be prompted for the location of the file If the Analysis Region shapefile specified in the analysis file no longer exists a message will appear to that effect The definitions of feature groups and output options will still be loaded Thus pre defined feature groups need not be re defined for each new Analysis Region 6 6 Help Selecting Help from the NCMS menu provides access to summary information on the application and its features The topics available are shown in Figure 6 11 Figure 6 11 Help topics Help I apt NOMS Help Contents Find Lick a book and then chock Open Or chek another tab such as Index 1 Diary ey F The NEMS Visualization and Analysis Sytem UL Meru Options ki Wia Spano 7 Define Regan Ki Deire Feature Groups 2 Retieve Anayiz E Help Ll E simal Land Lower 7 Step by Step Proc LL Land Cover Festus Detad 7 Antal and Liar Feah 28 National Countryside Monitoring Scheme Visualisation and Analysis System 7 Results The results of each analysis
7. this will give relatively low precision and the standard errors may be poorly defined but minimises loss of accuracy Minimise dependency on sample squares outwith the analysis region for estimates only Estimates are calculated using only inside squares where possible Standard errors are calculated using all squares in each stratum whether or not they fall within the defined region This means that relatively accurate estimates can be calculated with reduced standard errors This relies on the assumption that variability between squares is similar within and outwith the region Use all sample squares in strata Use this method to maximise precision This will be fairly safe if most squares are inside the region or if the strata that make up the bulk of the region are mostly contained within the region The greatest benefit is achieved when there are many outside squares that would otherwise be excluded from the analysis The diagnostic information should be used to assess the extent of dependency on outside squares A safer method should be used if there is a strong possibility of bias 25 National Countryside Monitoring Scheme Visualisation and Analysis System Define analysis options for each stratum individually This option displays the Advanced Analysis Options dialog Figure 6 9 and allows the analysis method to be varied between strata Figure 6 9 Advanced Analysis Options Advanced Analysis Options To change how results are
8. Countryside Monitoring Scheme NCMS is a major study of land cover change in Scotland Using aerial photography the study interpreted and mapped land cover for the late 1940s the early 1970s and the late 1980s for 467 sample squares across Scotland This enabled habitat change throughout Scotland to be quantified for the first time The NCMS Visualisation and Analysis System is an advanced ArcView GIS application which allows full interactive access to the NCMS dataset A flexible user interface permits visualisation and quantification of land cover change within each of the sample squares Land cover change within user defined geographical regions can be estimated through an extrapolation process The GIS tools are divided into two broad categories e Viewing tools e Analysis tools Viewing tools enable maps of land cover to be displayed for NCMS sample squares Key features include the following Selection of sample squares from a map or a list selection of the time period and features to be displayed examination and quantification of land cover changes a range of presentational layouts and creation of standard Views of selected squares suitable for further analysis Analysis tools enable the definition of geographical regions of interest for estimating land cover stock and change extrapolated from sample data Key features include the following e Regions of interest can be selected from existing coverages or by d
9. Extension to the default ArcView extensions folder For some network installations this folder may be inaccessible The setup program will warn you if it is unable to install the extension and you or your network administrator will have to manually copy the file ncms avx to the correct folder This file can be found in the folder called main on the CD 4 4 Un installing If you wish to keep any NCMS Analysis files that have been saved in the NCMS analysis folder first copy these to a safe location Then delete the folder that the application was installed to by default this will be C ncms Finally delete the NCMS extension ncms avx from the ArcView extensions folder National Countryside Monitoring Scheme Visualisation and Analysis System 4 5 Starting the Application 1 Start ArcView in the normal way 2 Select the Extensions option from the File menu 3 From the list of extensions tick NCMS and click OK A splash screen will be displayed while the extension loads A new View will be created called NCMS Main View Figure 4 1 This will have three themes e NCMS squares e Coastline e Stratification This View is a simplified version of the standard ArcView interface The only addition is an NCMS menu If any other window is activated such as a normal View ArcView will revert to its standard interface Figure 4 1 The NCMS Main View 2 National Countryside Monitoring Scheme Ele Edt View Theme NCMS Window Help Ok L
10. Visualisation and Analysis System Figure 5 9 Analysis Options 2 Analysis Options Analysis Metho f Minimize dependency on squares outwith analysis region Minimise dependency on squares outwith analysis region for estimates only Use all sample squares in strata Define analysis options for each stratum separately Special Case Certain features are automatically combined for some regions Separating them may produce invalid results see user quide Always separate grass and heather blanket mire Always separate built and transport coridor For now leave these options unchanged so that estimates are calculated to minimise dependency on squares outwith analysis region and click on Run A Summary Options window will then appear itemising the settings selected in the Outputs and Analysis Options dialogs Figure 5 10 Click on OK and you will prompted to save the settings in a file These analysis files can be retrieved for subsequent analyses for which the same or similar settings are required Specify a file name and a location for the file and click OK National Countryside Monitoring Scheme Visualisation and Analysis System Figure 5 10 Summary of selected options Summary Options Analysis options and the results to be calculated are summarised below Click OK to continue with the analysis or Cancel to return to the analysis options Analysis options Minimize depen
11. analysis The full extent of each stratum and its squares will be shown if outside 29 National Countryside Monitoring Scheme Visualisation and Analysis System Squares were used to calculate either the estimates or standard errors for that stratum In this way the geographical extent of the squares that are used can be visualised Heavy dependency on outside squares does not necessarily invalidate an analysis but results should be viewed with caution It is more satisfactory if squares inside the Region make the greatest contribution to the estimates but balanced against this is the benefit of improved precision if more squares are utilised The NCMS sampling design resulted in a number of small strata being undersampled see Appendix 2 7 2 Estimate reports Report tables provide estimates of stock net change or interchange between land cover features Separate report windows show estimates for areal and linear features and for interchanges Each estimate will have an associated 95 confidence interval or standard error depending on which was requested in the Output Options dialog As estimates are based on a sample there is uncertainty as to the true extent of each feature in the Analysis Region For change estimates there is uncertainty as to whether an estimated change represents a real change in the Region the change in the sample might not be reflected across the whole Region Confidence intervals define a range within w
12. are displayed as reports with each report in a separate window A table may occupy several pages within a report 7 1 Diagnostic Report For each analysis a diagnostic report summarises the utilisation of squares for the analysis and in particular the dependency on squares outside the Analysis Region The table shows the number of sample squares used to calculate estimates and standard errors and the area of each stratum in the Region Figure 7 1 Figure 7 1 Diagnostic report a T E E NATI IEEE IET aj ij Toi 1 ele e JES fars I am doi d Diagnostic Information nd T1 7 an 1 iti il 11 Ll 03 Triada M a 14 m L 7 1 ban Extent of strata and squares iied in th analysis The full extent is shown iE the Elim ales ar sLandard rrari mere calculated weing all the squares Estimates are calculated for each stratum and then summed to generate Region estimates Region estimates will be more dependent on squares outside the Region If estimates for the biggest strata in the Region are heavily dependent on outside squares This will be the case if a sizeable proportion of the squares used for these strata are outside the Region Strata that have a small proportion of their area in the Region are more likely to have required the use of outside squares but they will tend to contribute less to the overall estimate The map shows the Analysis Region together with the squares and strata used in the
13. further analysis and combination with other datasets calculate the area of each feature and select the size of layout for printing See Section 6 1 for further details 5 2 Estimating land cover change Estimation of land cover land cover change and interchange for a geographical region involves the following steps National Countryside Monitoring Scheme Visualisation and Analysis System Defining the region of interest the Analysis Region Specifying which results are required Defining feature groups optional Specifying how the results should be calculated Running the analysis ee IA To define the Analysis Region make sure the NCMS Main View is active and select Define Region from the NCMS menu The dialog shown in Figure 5 4 will appear Figure 5 4 Defining an Analysis Region 2 Definition of Analysis Region Please choose the method for defining the analysis region From an existing theme Cancel By drawing area s with mouse From current selection of active theme Regions can be defined from existing ArcView themes or by drawing the boundary on the NCMS Main View see Section 6 2 for full details For this exercise click on By drawing areas s with mouse and click OK A dialog will appear prompting you to save the Analysis Region theme Specify a file name and a location for the theme and click OK The drawing tools shown in Figure 5 5 will then be displayed Figure 5 5 Draw Analy
14. rh quera cover acu parlar E blunbet Bie bki ETE DMI raspa iir Lal id Saa Petit td Dp 54 rough gra13land ars pmd bil National Countryside Monitoring Scheme Visualisation and Analysis System Figure 5 3 The layout for the square selected in Figure 5 1 The report window shows the amount of change in the sample square from one feature to another between the 1940s and 1980s These changes are ordered by size no change indicates the area that was unchanged and outside cover includes areas within the sample square that were not interpreted usually because they fell outside a regional boundary or because insufficient aerial photography was available The report can be printed or saved for later use See Section 6 1 but for now simply close or minimise the report window to reveal the layout The layout shows the interpreted land cover for the sample square for c 1947 and c 1988 together with the areas that changed between these dates The largest 25 interchanges are colour coded and displayed in the legend ordered by size Smaller interchanges are grouped into other and areas that showed no change are left blank Figure 5 3 The size and grid reference of the square are shown in the bottom right of the layout with a map showing its location Other features of the View Square dialog enable you to select which dates to display display areal and or linear features create a standard View of the square to enable
15. sampling information to estimate land cover at different geographical scales Extent and change in extent can be estimated for each land cover type as well as the amount of change between pairs of areal features termed interchange A summary of the NCMS its method and main results can be found in the folder called ncms summary on the CD Double click on start htm or open this file in your web browser to begin National Countryside Monitoring Scheme Visualisation and Analysis System 3 Estimating land cover An understanding of how land cover is estimated from the NCMS sample squares is important for interpreting results for a region of interest Scotland was stratified through a classification of Landsat multi spectral scanner images To avoid difficulties of edge matching the strata were defined separately for each of the 12 former pre April 1996 Regional authorities and in some cases for each District in a Region From two to five strata were defined roughly corresponding to lowland upland intermediate and urban areas two upland strata were defined in Highland This resulted in a total of 80 strata overall Sample squares were initially 5km x 5km in Figure 3 1 NCMS sample size and a sample was selected randomly to square coverage cover 10 of each stratum A statistical review was subsequently able to optimise this design It resulted in the size of squares UR being reduced to 2 5km x 2 5km a quarter of the o
16. 7 s k sample square k 1 nj The intersections between regions and strata are illustrated by example below Region A Region B Intersection Area A Intersection Area Aj S l 100 0 S l gt 50 lo 200 strata S la 0 la 0 Sa la 0 l4 150 In this example strata S and S overlap with region A whilst S and S overlap with region B Estimates and standard errors Suppose we are interested in a measurement such as heather moorland area for a user defined region The observed areas of heather moorland within the sample squares are yj The method to be adopted is to form estimates for the region by summing estimates for the strata represented in the region The squares to be used for forming the estimate will either be those within the intersection of the region with strata S or all squares within S The number of squares used is nj 42 National Countryside Monitoring Scheme Visualisation and Analysis System An estimate Y of the total area or length for each non empty l is required An intuitive approach is to use the ratio estimator 2 k l A pa j J 2205 k 1 J The total area or length for the region can then be estimated by Y DY j j l with corresponding variance njAj l aj A 5 nj 2 var Y 2 nj 1 a D On i Kd jp j 1 k 1 where nj gt Y jk ni k l aT k 1 S k 1 and A Aj if only squares in lj are used or the area of the stratum if all squar
17. ESTIMA TNG LAND COVER CHANGE e EEEE la e 9 6 APPLICATION REFERENCE ui il 17 O30 VIEWSOUARE saetas 17 0 2 DEPINE REGION ct o a dla e e e de a ds o 20 0 9 DERNE PEATURE AROS e la ear a enero o ro aar 22 ode UNE WAN ADS lesa hg Sencha bc estu ceca do tus oU aeu nem tof donas e Cone atado scudo nidos 23 ES EA i t A ae tee dud 23 042 SAO A ei ene oe ee 24 Ov D gt OHIO TRE GN OLY SIS ia 27 O ra UN OLESTIE MESA Sg ae le ee 28 Co HEL ran tear cer cists os ca esa ecto e tie ac odd e esc rete IM cedat pa Pei ecd 28 Je RESULTS vaia ici 29 PI SIDENGINOS TIO REPORT e al denon Eo 29 T ESTIMATE REO ds O M CE 30 REFERENCES a did 32 APPENDIX 1 THE NCMS LAND COVER FEATURES c cccsccssccccecccccccscccscccccccccceces 33 APPENDIX 2 DATASET ANOMA LEDS 60 isos 37 Changes LOINC classic oe 34 SAMPINTIS MES cdo el 37 Gaps um The SANN ECOIN A A A AA A A EE 39 BIS RA RA AS AL AA A IS A A ES AL ETA 39 APPENDIX 3 ESTIMATION METHODS unidad 40 APPENDIX 4 CALCULATION OF RESULTS eee ei e e esvescos ere eoe ss ecu esee oe oaa deco sere se ee os ren es epe Pr uos 42 VR SI Ia mE 42 Estimates and standard E id 42 EA A Te EVO ASS A AS is 44 APPENDIX 5 SQUARE AND STRATA CODES eee eee ee eene eee ee oett eene eese etes ee eee eee etes eas 45 APPENDIX 6 FEATURE CODES aii as 46 APPENDIX 7 METADA DA toos 47 National Countryside Monitoring Scheme Visualisation and Analysis System 1 Introduction The National
18. an be calculated by summing the separate estimates standard errors and confidence intervals can only be found by analysing the group as a whole The dialog for creating groups Figure 6 6 can either be accessed from the Define Feature Groups option on the NCMS menu or from the Outputs dialog see Section 6 4 1 Figure 6 6 Grouping features for analysis Areal Groups To group features for analysis create a Mew Group and add features from the list Groupe Change to Linear Groups Features Features in group arable New G bare ground lia blanket mire grass blanket mire heather bracken broadleaved plantation broadleawed woodland built canals Delete Group peel 22 National Countryside Monitoring Scheme Visualisation and Analysis System Grouping features together involves creating a group such as woodland and then allocating features to that group 1 2 6 To create a new group click on the New Group button and enter a name for the group Select a feature and use the 1 button to add it to the group Repeat until all the required features are in the group You can use the button to remove features The Groups drop down list can be used to move between different groups Click on the Delete Group button to completely remove a group all features in that group will be returned to the main list The Change to linear area
19. any Analysis Regions but the diagnostic information provided with each analysis will show how extensive the strata are in the Region 38 National Countryside Monitoring Scheme Visualisation and Analysis System Characteristics of strata with no sample squares Strata NCMS district Type Total area km sis1 Strathclyde Islands Agricultural 84 sis2 Strathclyde Islands Moorland 126 sna1 North Argyll Agricultural 46 sna4 North Argyll Urban 6 Gaps in the stratification Certain small islands such as Rum were never included in the NCMS stratification and therefore not sampled These areas are white on the stratification coverage Whenever an Analysis Region includes such islands the results will be calculated for the stratified part of the Region Coastline The coastline presented in the NCMS Main View is relatively crude having been derived from the original NCMS stratification rather than from an Ordnance Survey dataset Slight discrepancies with other coastline datasets should not be significant for most purposes 39 National Countryside Monitoring Scheme Visualisation and Analysis System Appendix 3 Estimation methods Three methods are provided for calculating estimates and standard errors for each Stratum in a region of interest The method of choice will depend on whether maximising sample size and hence precision is more important than minimising the dependency on squares outside the region Method A Minimi
20. calculated select a stratum and choose which squares to use for estimates and standard errors Ma of Use all Squares Squares for Strata In Total Estimate SE Use all squares in strata to calculate Estimate Standard Error To change the squares that are used for each stratum click on the stratum and tick the Estimate box to use all squares for the estimate which will then register true Tick the Standard Error box to use all squares for standard errors and hence confidence intervals If these boxes are not ticked only squares inside the Analysis Region will be used Only valid methods can be specified for example it is not possible to calculate estimates for a stratum using only inside squares if there are no inside squares This option is used to fine tune an analysis by increasing the number of squares used for a stratum that may otherwise produce unreliable estimates For most strata the number of inside squares may be adequate but for a few it may be desirable to use outside squares either to improve precision or to improve the validity of the estimate For example a stratum may have only one square in the Region and four outside An estimate based on the one square may not be representative of the land cover for that stratum whereas the five squares combined will be more so Similarly two squares are enough to calculate a standard error but one calculated from five squares is likely to have greater validity an
21. cisively rough or smooth the sward composition of intermediate grassland may appear to have been modified to a greater degree than rough grassland by land management practices such as the application of fertiliser or herbicide heavy grazing pressure or land drainage smooth grassland Heavily modified by the application of fertilisers and or herbicides and may have been re seeded note that temporary grassland ley within an arable rotation was classed as arable distinguished from smooth grassland by contextual detail Mire blanket mire heather Landscape form and vegetation cover together with surface morphology were used to infer the sub surface presence of wet acidic peat under heather If drained heather dominated mire was classified as heather moorland When stripped of Its surface vegetation it was classified as bare ground or quarry where there was clear evidence of peat extraction blanket mire grass Landscape form and vegetation cover together with surface morphology to infer the sub surface presence of wet acidic peat under rough grass lowland mire Generally related to dome formations of raised bog but includes unwooded fens Heather Moorland Dwart shrubs or regenerating burnt patches exceed 50 of the ground cover 33 National Countryside Monitoring Scheme Visualisation and Analysis System Arable Without differentiating between crops the classification includes arable crops rotational grasslan
22. ck on the Define Groups button and the dialog shown in Figure 5 8 will appear Click on New Group and you will be prompted for a name for the group For example type in mire and click on OK You can add features to National Countryside Monitoring Scheme Visualisation and Analysis System this group from the list on the left Click on blanket mire grass and then on the button to place the feature in the group Repeat for blanket mire heather and lowland mire Finally click on OK to return to the Outputs dialog The three mire features are now grouped together and a combined estimated will be calculated in subsequent analyses Further information on combining groups can be found in Section 6 3 Figure 5 8 Grouping features for analysis 2 Areal Groups To group features for analysis create a Mew Group and add features from the list Groupe Change to Linear Groups Features Features in group Mew Group arable bare ground blanket mire grass blanket mire heather bracken ES braadleaved plantation braadleaved woodland built canals Click on OK in the Outputs dialog to bring up the Analysis Options dialog Figure 5 9 This allows you to affect how the estimates are calculated and in particular the extent to which sample squares outwith the Analysis Region are used see Section 6 4 2 for details National Countryside Monitoring Scheme
23. d give a more precise estimate Special Cases As discussed in Section 6 3 blanket mire grass and blanket mire heather were not differentiated in all strata for the 1940s and 1970s Similarly for built and transport corridor see Appendix 2 for details Thus if the Analysis Region includes any such strata these features will be automatically combined However it may be evident from the diagnostic report that such strata only account for a small proportion of the Region In such cases you can tick either or both of the check 26 National Countryside Monitoring Scheme Visualisation and Analysis System boxes to force these features to be kept separate see Appendix 5 for a list strata codes Use these options with caution If the strata for which the features were not differentiated occupy a substantial proportion of the Region misleading results may be given For example all built and transport corridor was coded as built in Grampian and Lothian in the 1940s and 1970s In this case results comparing the 1940s and 1980s may suggest an apparent sharp increase in the amount of transport corridor if the features are kept separate Finally the Analysis Options dialog can be used to select whether or not feature groups see 6 3 are to be used in the analysis Accordingly the analysis can be based on all individual features or on any groupings that have been defined 6 4 3 Running the analysis Clicking th
24. d ley and horticulture Woodland broadleaved woodland Broadleaved crowns account for more than 50 of the area and coniferous cover is less than 25 Tree height is taller than 5 metres mixed woodland Trees greater than 5 metres tall natural or planted in appearance and composed of more than 25 broadleaved and more than 25 coniferous tree cover broadleaved plantation Even aged stands of more than 50 broadleaved trees normally not native to the site and dominated by one species parkland May be a coniferous or broadleaved grouping of at least 10 trees providing 10 to 50 cover coniferous woodland Irregular tree cover in excess of 50 and less than 25 broadleaved cover young plantation Regularly planted coniferous or broadleaved but negligible in extent trees of up to 3m in height coniferous plantation Regularly planted conifers exceeding 50 of trees present with broadleaved cover less than 25 felled woodland Areas in which at least 50 of woodland had been recently felled Fresh Water lochs Irregular shaped expanse of water without evidence of impounding reservoirs Expanse of water with evidence of impounding rivers Flowing water of greater than 10m in width without evidence of canalisation streams Flowing water of less than 10m in width without evidence of canalisation 34 National Countryside Monitoring Scheme Visualisation and Analysis System canals Water course greater than 10m w
25. dency on squares outwith analysis region Use Defined Groups 398 Confidence Intervals Areal results for 1940371 980s 1 940s to 1980s Linear resulte for 1940 19580 1940 to 1980s The system will then proceed with calculation of the estimates This can take some moments depending on the options chosen and the size of the Analysis Region Once complete several report windows will appear On top will be the diagnostic report which for the Western Isles Analysis Region should look like that in Figure 5 11 This report summarises the dependency of the results on squares outwith the Analysis Region see Section 7 1 for a full description In this example the Analysis Region contains only two NCMS sampling strata and all the sample squares in these strata fall within the Region Thus there is no dependency on outside squares This will rarely be the case and for most Analysis Regions some outside squares will have to be used National Countryside Monitoring Scheme Visualisation and Analysis System Figure 5 11 Diagnostic report for the Western Isles E Digna dm mmm FECE aja tot 0E E E x Tm 1 T Diagnostic Information Ta kaia E L ia ii 3013 109 0 Exirr af strate aed eq ceed in ihe xzxlyzia fhe full ria im chown Lf the rta or siandard ARTO were ralceulabad mama all he are Close or minimise this window to reveal the estimates for linear features which should be similar to that shown in Figur
26. dernisation and expansion of farming and forestry to around 1973 when the UK joined the European Economic Communities and then to 1988 as environmental balance in rural policies was receiving greater recognition It was impracticable to interpret photography for the whole of Scotland Instead a stratified random sample representing 7 596 of Scotland s land area was developed to estimate the extent of land cover features and changes through time otratification was performed on a regional basis by the classification of Landsat multi spectral scanner images Each stratum was sampled to select photography for interpretation The sample size was determined by a need to detect national changes of 1096 or more in the major features with 95 confidence Land cover was classified in terms of 31 areal features such as heather moorland and five linear features such as hedgerow see Appendix 1 Each sample square there were 467 in all of 5km x 5km or 2 5km x 2 5km in size was mapped at a scale of 1 10 000 The minimum mapping resolution was around O 1ha for areal features and 30m for linear features Land cover maps for each sample square were digitised for processing on a Geographical Information System GIS This enabled the land cover in each sample square to be calculated for each time period Overlay analyses performed the computation of land cover change between time periods Statistical software allowed GIS outputs to be combined with
27. ds and includes unquarried inland cliff unvegetated rock scree and limestone pavement Due to the perspective of the photography steeply inclined or vertical cliff face such as coastal cliff are not represented bare ground Land not covered by vegetation which does not fall into any other category It represents a transition phase and includes erosional features such as exposed gravel or soil in upland areas but not temporarily bare arable 35 National Countryside Monitoring Scheme Visualisation and Analysis System Bracken and Scrub bracken At least 50 bracken cover low scrub Trees or shrubs which are less than 3m in height with a range of canopy sizes and heights tall scrub Dense canopy of trees or shrubs of 3m to 5m in height Hedgerows and Trees hedgerow Shrubs and trees less than four metres in height and five metres in width classified as continuous if gaps are less than 10 metres wide treeline Often encountered along field margins a minimum of 3 trees of greater than 4 metres in height and less than 2 canopy widths apart 36 National Countryside Monitoring Scheme Visualisation and Analysis System Appendix 2 Dataset anomalies For certain areas of Scotland anomalies in the way sample squares were selected or in the way the classification was applied may affect the interpretation of results In the vast majority of cases these will not be important but the description below should enable an assess
28. e Run button on the analysis options dialog presents a summary of the selected output and analysis options Figure 6 10 Figure 6 10 Summary of analysis options Summary Options Analysis options and the results to be calculated are summarised below Click OK to continue with the analysis or Cancel to return to the analysis options Analysis options Minimize dependency on squares outwith analysis region Use Defined Groups 398 Confidence Intervals Areal results for 194071 980s 1 940s to 1980s Linear resulte for 1940 19580 1940 to 1980s Check the settings are as required and click OK to continue with the analysis or Cancel to return to the Analysis Options dialog If you continue you will be prompted to save these settings in an analysis file The analysis file can be opened at a later date to repeat a particular analysis see Section 6 5 Specify a file name and a location for the file and click OK 27 National Countryside Monitoring Scheme Visualisation and Analysis System The system will then proceed with calculation of the estimates This can take some moments depending on the options chosen and the size of the Analysis Region 6 5 Retrieve Analysis Key settings used in the analysis are saved in an analysis file see 6 4 3 These can be retrieved to repeat or modify an analysis i e without having to define the analysis again from scratch It can also be used to perform the same analysis on a
29. e 5 12 The estimated length of each linear feature is given for the 1940s and 1980s as well as changes between these dates The lower and upper figures are the 95 confidence limits around each estimate A guide to the interpretation of results can be found in Section 7 2 Figure 5 12 Linear feature estimates for the Western Isles H M Laa rimai z l ef oe obie aian a res ua eat Linear Estimates kim Minimising or closing the linear estimates report will reveal a similar report containing estimates and confidence intervals for areal features This table should contain estimates for mire the feature group defined earlier National Countryside Monitoring Scheme Visualisation and Analysis System Finally you can save results in a variety of file formats for further analysis or presentation see Section 6 1 National Countryside Monitoring Scheme Visualisation and Analysis System 6 Application Reference The NCMS menu has the following options which are described below Menu option View Square Define Region Define Feature Groups New Analysis Retrieve Analysis Help 6 1 View square Description Opens the NCMS Square Viewer dialog allowing squares to be selected and displayed Allows the user to create an Analysis Region the first step in the estimation process Allows features to be combined into groups prior to analysis Allows the user to specify and perform an anal
30. e click on the print button above the report To save the table in a spreadsheet or other format click on the export button Ei This brings up the dialog shown in Figure 6 2 Select a Format to save the report in and a Destination usually Disk file Click OK and you will be prompted for the name and location of the file to be saved Figure 6 2 Exporting reports Format Ls urn ALa Cancel Export Destination Disk file This report is overwritten each time a new square is viewed If a View of the selected square has been requested each date and interchange will be displayed as separate theme Figure 6 3 An example of a layout is shown in Figure 5 3 Finally if you wish to highlight particular features or groups of features in a square by selecting them you will need to construct a query by selecting the theme you wish to query in the square s View and then selecting Query from the Themes menu The underlying NCMS dataset uses short letter codes to identify land cover features and these have to be used in queries instead of the feature names The table in Appendix 6 shows the code used for each feature National Countryside Monitoring Scheme Visualisation and Analysis System Figure 6 3 The View of the square selected in Figure 6 1 d gii eee Hia Bb 40 807 EM ng gees le naar eee 082323 hanh a te bc dll id S d gata reeled eg graaalared 223 07 LIEL CTET TEIT ice bobathai 10d birki r
31. e feature names themselves In normal usage these codes will remain hidden but if for example you wish to construct a query to select particular features in a square you will have to use feature codes The table below shows the code used for each feature Feature name Feature code Feature name Feature code Areal Linear arable A ditches DITCH bare ground BG hedgerows HEDGE blanket mire grass BMG lines of trees TREE blanket mire heather BM streams STR bracken BR tracks TRACK broadleaved plantation BP broadleaved woodland BW outside cover OC built B canals C coniferous plantation CP coniferous woodland CW felled woodland FW heather moorland HM intermediate grassland IG lochs L low scrub LS lowland mire LM marginal inundation MI mixed woodland MW parkland P quarry Q recreation R reservoirs RE rivers RI rock RK rough grassland RG smooth grassland SG tall scrub TS transport corridor TC wet ground WG young plantation YP As detailed in Appendix 2 built and transport corridor and heather dominated and grass dominated blanket mire were not distinguished in certain Regions in the 1940s and 1970s interpretation In these cases the code B was used for both built and transport corridor and BM was used for both heather dominated and grass dominated blanket mire 46 National Countryside Monitoring Scheme Visualisation and Analysis System Appendix 7 Metadata Title Alternative Title Originato
32. elineating areas with the mouse control over extrapolation methodology habitat features can be grouped according to user interest analysis reports can be saved for further analysis or presentation and analysis settings can be saved for later re use and adjustment The system is straightforward to use Nonetheless a basic knowledge of ArcView would be helpful and to make best use of the analysis tools and results an understanding of statistical sampling and estimation is recommended National Countryside Monitoring Scheme Visualisation and Analysis System 2 The National Countryside Monitoring Scheme Looking back over the latter half of the twentieth century considerable changes have become evident in Scotland s urban and rural environments Some changes such as urban expansion and afforestation have been striking More subtle and cumulative changes such as in the structure of farmland or the extent and condition of moorland may be less obvious Yet they are no less relevant to the visual appearance of the countryside to its wildlife and to opportunities afforded for outdoor recreation and enjoyment Archive air photography has made it possible to look back over this period to investigate and quantify change its magnitude rate and geographical variation A land cover audit known as the National Countryside Monitoring Scheme NCMS sampled photography from around 1947 as the imperative for economic growth promoted the mo
33. erval or standard error for an estimate these values are left blank 31 National Countryside Monitoring Scheme Visualisation and Analysis System References Mackey E C Shewry M C amp Tudor G J 1998 Land Cover Change Scotland from the 1940s to the 1980s The Stationery Office Edinburgh Tudor G J Shewry M C Mackey E C Elston D A amp Underwood F M 1999 Land Cover Change in Scotland The Methodology of the National Countryside Monitoring Scheme Scoitish Natural Heritage Research Survey and Monitoring Report No 127 Scottish Natural Heritage Perth Elston D A Gauld J H Miller J A Shewry M C amp Underwood F M 1999 The National Countryside Monitoring Scheme Accuracy Assessment Scottish Natural Heritage Research Survey amp Monitoring Report No 133 Scottish Natural Heritage Perth 32 National Countryside Monitoring Scheme Visualisation and Analysis System Appendix 1 The NCMS Land Cover Features The following descriptions provide a guide to how NCMS features were interpreted from aerial photographs Grassland rough grassland Rank or tussocky grassland which may appear to have been drained grazed mown or treated with farm manure but not so improved by fertiliser or herbicides as to have altered the sward composition greatly Associated with unenclosed upland sites lowland sites with poor access or wet areas and may include roadside verges intermediate grassland Not de
34. es in S are used R Note that the land cover estimate for stratum could be calculated using only squares in and the variance calculated using all squares Thus each strata can be treated differently and the calculation of variances treated differently to the calculation of the estimates themselves Note that the estimate can only be calculated using just the squares in if there is at least one sample square in Similarly the variance can only be calculated in this way if there are at least two squares in Estimates for land cover change are calculated in exactly the same way substituting the interpreted change in the feature within square k for y Similarly estimates of interchange between two features are calculated by substituting the interpreted interchange between the features within square k for yi To calculate the standard error of the estimate simply take the square root of the variance 43 National Countryside Monitoring Scheme Visualisation and Analysis System Confidence intervals 95 confidence limits for estimates or land cover stock or interchange are calculated assuming a log normal distribution lower limit A lt gt upper limit Y K where K exp 1 96 var log Y and Y var log Y log 1 He y Y This gives asymmetric confidence intervals with the lower limit closer to the estimate than the upper limit The upper and lower 95 confidence limits for es
35. hich the true value lies with the specified level of confidence e For estimates of extent or interchange confidence intervals define a range for the true extent or interchange in the Analysis Region e For estimates of change if the range of the confidence interval is all positive or all negative then it is likely that the change is real i e if there had been no change in the Region there is less than a 5 chance of the estimated change being as large The use of 95 rather than say 90 confidence intervals is to some extent arbitrary but it ensures that not many changes are interpreted as being real when in fact no change took place Standard errors provide an alternative measure of uncertainty or precision ina feature s extent or change As standard errors will be larger for larger features one way of comparing the precision between features is to calculate the coefficient of variation which is simply the standard error divided by the estimate Standard errors can also be used to calculate confidence intervals using the formulae presented in Appendix 4 Other sources of error are inevitable in a study as complex as the NCMS These are more difficult to quantify than those due to sampling but are worth bearing in mind Errors can occur when interpreting features from aerial photography and when digitising boundaries These errors are not thought to result in substantial bias in the majority of cases further information can be f
36. idth which has been artificially confined to flow ina certain direction ditches Water course less than 10m width which has been artificially confined to flow in a certain direction marginal inundation Included swamp or fen margins typical of an open water transition the banks of ponds and ditches subject to periodic inundation and the draw down zones of reservoirs wet ground Small areas of wet land such as wet areas in a pasture field or flushes in upland areas often denoted by the presence of the rush Juncus without evidence of peat formation Built and Bare Ground built land Urban areas including buildings roads gardens parks and golf courses within the urban boundary and buildings outside urban areas recreation Land in the countryside normally adjacent to urban areas which was in formal recreation use Includes sports fields playing fields golf courses camping and caravan sites ski runs and racing circuits transport corridor Railways and surfaced roads greater than 3 metres in width which occurred outside built up areas greater than one building deep on either side of the road including overbridges carriageways hard shoulders and other unvegetated roadway features tracks Non transient routeways of up to 3 metres width showing signs of use by wheeled vehicles quarry Hard rock excavations sand and gravel pits open cast mines and unvegetated spoil heaps rock Associated mostly with the uplan
37. ithin themselves The validity of this assumption can be evaluated from diagnostic information accompanying each analysis For regions where this assumption proves questionable the results of the analysis should be regarded as indicative only National Countryside Monitoring Scheme Visualisation and Analysis System 4 Getting Started 4 1 Pre Requisites Software The NCMS application is an Extension for ArcView GIS 3 2 running within a Windows operating system Crystal Reports supplied with ArcView must be installed Hardware The application uses some very large datasets requiring at least a Pentium 200 processor and 64mb of RAM Speed can be greatly improved if the data are read from the hard drive of the PC rather than the CD ROM If the datasets are kept on the CD ROM about 3mb of disk space is required for installation If the data are to be copied to the hard drive a total of about 650mb of disk space is required 4 2 Installation Insert the CD and if autorun is enabled the installation program should start automatically Otherwise run the program X setup exe where X is the drive letter of your CD ROM drive and follow the on screen instructions A number of folders will be created to hold the NCMS data reports analysis files online help and temporary files You will be given the option to install the datasets onto the PC s hard drive 4 3 Network Installation The installation program will attempt to copy the NCMS
38. l groups button is used to switch between the two structural types Groups cannot contain both areal and linear features Click OK when finished Two special cases 1 A distinction between blanket mire grass and blanket mire heather was not made in the strata in Borders Dumfries amp Galloway Grampian Lothian Orkney Islands or Shetland Islands for the 1940s and 1970s By default these features will be grouped together as blanket mire if the Analysis Region includes any such strata If they have been placed in separate groups they will be removed automatically from those groups and grouped as blanket mire for analysis Similarly built and transport corridor were not identified separately in the Grampian and Lothian strata for the 1940s and 1970s They are therefore treated in a similar way to blanket mire 6 4 New Analysis 6 4 1 Specifying outputs The Outputs dialog is used to specify the dates and features for which estimates are required This dialog appears once a new Analysis Region is defined or it can be accessed by selecting New Analysis from the NCMS menu Figure 6 7 23 National Countryside Monitoring Scheme Visualisation and Analysis System Figure 6 7 Outputs dialog 2 Outputs Stock e 1940 BB e 1980s Net Chang scs T orsus T P40s 80s Feature e Amal wv Linear Interchang 40s 70s 7 s 80s M 40s 80s I 95 Confidence Intervals D Standard Error
39. lect whether to display areal and or linear features Format Data can be displayed in a standard ArcView View and or a presentational Layout Calculated Areas If this box is ticked then the map legends will show the extent of each feature in the square for each date selected Layout Template Select the required size of Layout for printing using the drop down list Note that a distinction between blanket mire grass and blanket mire heather was not made in Borders Dumfries Galloway Grampian Lothian Orkney Islands or Shetland Islands for the 1940s and 1970s Most of the blanket mire in these Regions will have been coded as blanket mire heather for these dates Similarly built and transport corridor were not identified separately in Grampian and Lothian National Countryside Monitoring Scheme Visualisation and Analysis System for the 1940s and 1970s and transport corridor will have been coded as built for these dates Thus an apparent sharp increase in the area of transport corridor between the 1940s and 1980s say is likely to be due to built land being reclassified as transport corridor for the 1980s maps Interchanges between these features are also likely to be misleading If a change map is selected a separate report window will appear displaying the full list of interchanges for an example see Figure 5 2 To switch between pages use the arrow keys S HH To print a tabl
40. ment of their likely effect on particular analyses Changes to the classification As the NCMS study progressed it became apparent that changes to the feature classification were required to enable study of particular issues Two features were sub divided after the 1940s and 1970s interpretation had been completed in some regions e No distinction was made between built and transport corridor in Grampian and Lothian Regions for the 1940s and 1970s interpretation The distinction was made elsewhere and in all Regions in the 1980s interpretation For simplicity these features have therefore been combined whenever the Analysis Region contains Lothian or Grampian strata A warning to this effect will appear before the analysis is performed e No distinction was made between heather dominated and grass dominated blanket mire in Borders Dumfries amp Galloway Grampian Lothian Orkney Islands or Shetland Islands for the 1940s and 1970s interpretation Again for simplicity these features have been combined whenever the Analysis Region contains a stratum from these regions Sampling issues Strata with only one sample square At the outset of the study a 10 sample was selected within each stratum In two small strata this resulted in only one sample square being selected This is insufficient to enable estimation of the standard error for these strata In the analysis these standard errors have been set to zero which res
41. of the sample squares used lie outside the Region This can be checked visually by examining the Analysis Region together with the stratification and sample squares in the NCMS Main View Selecting Define Region from the NCMS menu provides three methods for defining an Analysis Region e From an existing theme selects an area or areas from an existing coverage e By drawing area s with mouse allows you to define your area of interest using ArcView s drawing tools e From current selection of active theme uses the currently selected area in the NCMS Main View 20 National Countryside Monitoring Scheme Visualisation and Analysis System a From an existing theme selecting this option will display the dialog shown in Figure 6 4 Figure 6 4 From an existing theme dialog 2 Define Region Make Analysis Region theme fram theme New Use the selection tools below to select the features to be made into the Analysis region Click OF when you have made your selection Selection Tool OK Cancel If you have already added an ArcView theme from which the Analysis Region is to be defined to the NCMS Main View you can select it from the drop down list Otherwise click on the New button and select the file containing this theme from the Add theme dialog that appears Next use the selection tools to select the polygons in this theme that define the Analysis Region
42. ound in the references provided Figure 7 2 shows a report of area estimates for the 1940s and 1980s and net change between the1940s and 1980s with 95 confidence levels 30 National Countryside Monitoring Scheme Visualisation and Analysis System Figure 7 2 Example of an estimates report oe HIS Aral Lain aja tt s jaja mn mx mn Area Estimates sq km ac I int 7 ma ais T T Ln tu poe E amp E i j a i bh ins i E i A n i E 1 bruta I I i L i 3 y baltinarpor oo i E d id Fi M i J J ruri 1 E r F a D a i a ei ikri a i bere il I An Iki T T mm Hl LI LITERAS BAD ud 1 Ift ra i T Lu i E LL Ex la i T L 1 j Lor er E l n I iz z La E i bra imde a La i T Ir I J A Ca eee jos E E 1 E a I T pond Datta i L l i I z i i i i E I L 7 EEM is a a j l a i rpg A i E l L 1 mau I E al i F i a i i ik E 5 E a E Iz agb paar E TL ra ml Th HT um 111 kl tots E gene bree HI Hej n 1M E am ul a Hal Il ees 3 T l j 3l I m 3 i uel trigal j 11 H i F i il A i mai a q Exi 1 ER Hi E SL ni Tatili 1 142 1 793 In this example the confidence interval for the change in rough grassland is all negative and so we can be reasonably confident that the change is real The same cannot be said for smooth grassland as its confidence interval ranges from 121 to 246 If insufficient data are available to calculate a confidence int
43. pendix 3 for further detail 24 National Countryside Monitoring Scheme Visualisation and Analysis System Special Cases permits the automatic grouping of certain features to be overridden see Section 6 3 This will only be required in exceptional circumstances Figure 6 8 Analysis Options 2 Analysis Options Analysis Metha f Minimize dependency on squares outwith analysis region Minimise dependency on squares outwith analysis region for estimates only Use all sample squares in strata Define analysis options for each stratum separately Special Case Certain features are automatically combined for some regions Separating them may produce invalid results see user quide Always separate grass and heather blanket mire Always separate built and transport coridor Analysis Method The more sample squares that are used in an analysis the more precise will be the resulting estimates Hence confidence intervals will be narrower Maximising the sample size usually means making use of data from squares outwith the region This may reduce the accuracy of the results if the outside squares are not representative of land cover within the region The Analysis Options dialog permits selection of the extent to which outside squares are used The options are as follows Minimise dependency on sample squares outwith analysis region Outside squares are used only when unavoidable As the safest option
44. r Abstract Access Constraints Data capture period Frequency of update Keywords Geographical Extent system of Spatial Referencing by Co ordinates Contact title of supplier Postal address of supplier Email address National Countryside Monitoring Scheme NCMS Scottish Natural Heritage Study of land cover change in Scotland between c 1947 c 1973 and c 1988 Aerial photography was interpreted for a stratified a random sample of 5km by 5km squares later 2 5km by 2 5km and digitised and processed on a GIS otatistical software allowed sample square data to be extrapolated to provide estimates of extent change and interchange for a geographical region of interest Land cover was interpreted in terms of 31 areal land cover features and five linear features For each of the 467 sample squares the interpreted features were mapped at a scale of 1 10 000 using a minimum mapping resolution of around 0 1 hectares for areal features and 30 metres for linear features oee licence agreement C 1947 c 1973 c 1988 No update planned conditional on comprehensive air photo coverage land cover change ocotland British National Grid NCMS support ocottish Natural Heritage 2 Anderson Place Edinburgh EH6 5NP ncms snh gov uk 47
45. ra harta br wl mita gare T8 bth ul ote he hie a laad Cif zi lara rera grarpfhlar amp qk erg beer cr Es EN ci giagg Lan ee eT i in Grit cave blink mus hari ar a ee ar eve Ll erie es clo pg E bth ul in en Fd 25 aab re arlasdiblanh sz mire granr i121 Wee hehe ered 1 A T hat meat landi cee 1 rc mia e prat 15 al ee ad ir 5 1 j AA erar cid n bunkri mis heshe eee rm idh ari A mie gia Q2 73 a te cere EFFES kd us s ati O 6 2 Define Region An Analysis Region can be defined by selecting or drawing a boundary in the NCMS Main View This Region can be represented by a single contiguous area or a series of geographically separate areas Although estimates will be calculated for almost any Analysis Region they will not be meaningful if the region is either very small or consists of NCMS strata whose sample squares lie mainly outside the Region As a rule of thumb Regions smaller than the area of Fife about 1 500 km are likely to give estimates with wide confidence intervals or will contain too few sample squares to enable reliable estimation The exception to this rule is that if the Analysis Region corresponds with a former local authority District it will usually contain sufficient squares Large regions may nevertheless present problems if results are heavily dependent on outside squares An examination of the diagnostic information should make it clear if there are any sizeable strata for which most
46. riginal size and a fixed sample size of five squares per stratum being selected T Dd The intensity of sampling therefore varies F between strata G ye E Estimates for a geographical region are i S ES m formed by calculating estimates for each de a stratum in the region and adding them E PET together Stratum estimates are formed by Em o un extrapolating the data from the squares in T ORQUR IAS each stratum Standard errors and confidence intervals are calculated to provide measures of uncertainty due to sampling In a study as complex as the NCMS there are inevitably sources of error e g when interpreting features from aerial photography and when digitising boundaries These errors are not thought to result in substantial bias in most cases see the references for further information User defined regions of interest termed Analysis Regions are likely to cover several NCMS strata either partially or completely Some may be poorly represented by NCMS squares within the region or in extreme cases contain no sample squares at all In such cases it is necessary to include squares falling outside the Analysis Region when calculating estimates Therefore estimation relies upon an assumption that strata are sufficiently uniform for squares outwith the region to be representative of land cover within the region In many cases this is likely to be valid as strata are generally small and reasonably homogeneous or similar w
47. scottish Natural Heritage Land Cover Change in Scotland The National Countryside Monitoring Scheme Visualisation and Analysis System User Guide SCOTTISH NATURAL HERITAGE Baal E GeoWise N GIS Consultants National Countryside Monitoring Scheme Visualisation and Analysis System Use of this software is subject to a licence agreement The full licence can be found in the file nemslicence pdf on the CD ROM In addition a summary highlighting the main terms of the licence can be found in nemsfaq paf National Countryside Monitoring Scheme Visualisation and Analysis System Contents hs INTRODUCTION il A aaa aE 1 2 THE NATIONAL COUNTRYSIDE MONITORING SCHEME sesessscsesecsesesccseseosesesecseseesese 2 S ESTIMATING LAND COV ER iia o eov ene eei seca eene eyR E e i o Ce coro a ev ee aa eve pus 3 As GETTING STARTED aiedi s seceeee ovest e veces ee rae esee eoe e e gara aa ou eG Y cue oe a Ne de VOV eoe eue ro Pau Puva ceo vu ee voee ua Er oUe 4 b OPRESREOUISEITBS 3 20 25 5 onmoscie es bio Ete epre cet e a ci Ste voPu Edu atch bevaset pam ML tecaramsscwanaue 4 A2 E A ce ed n E n aun EE ea M EE DUT eee 4 43 INEIWORKINSEPAELEATION i e rinlct beotres veo lu d s pee d beue dia 4 ZB EMEN US RESI V B B SO NR IR E AA T EA 4 AS STARTING THE APPLICATION id 5 Se AUTORIAL a aa Ta aa 7 5 1 DISPLAYING LAND COVER IN A SAMPLE SQUARE ccssccseccssccsccsceusccsccusccsccesccescesscuscescesscescenses 7 32
48. se dependency on sample squares outwith the region possible if the stratum has two of more sample squares in the region Estimates and standard errors for the stratum are calculated only from sample squares that fall within the region Method B Minimise dependency on sample squares outwith the analysis region for estimates only possible if the stratum has one or more sample squares in the region Estimates are based only on sample squares that fall within the region but standard errors use all the sample squares in the stratum This increases the reliability of standard errors by increasing the sample size for their calculation but assumes that the variability between squares is similar within and outwith the region Method C Use all squares in the stratum can always be used Estimates and standard errors are based on all squares in the stratum regardless of whether they are inside the region This increases the precision of the estimates by maximising the sample size available for estimation but makes the strongest assumptions Not only is the assumption for Method B required but also that the land cover in the strata is sufficiently homogeneous that squares outside the Region can be considered representative of land cover inside the Region The validity of these assumptions is difficult to assess but diagnostic information provided with the results of each analysis can help to assess the extent of the dependency of the analysis on ou
49. sis Region dialog 2 Draw Analysis Region Draw the boundary of the areale using the tools below Click OK when Finished i Cancel Try using the polygon tool selected in Figure 5 5 to draw an Analysis Region enclosing the Western Isles as illustrated in Figure 5 6 Remember to double click to define the last vertex of the polygon Click on OK in the Draw Analysis Region dialog and the Analysis Region theme will be added to the NCMS Main View National Countryside Monitoring Scheme Visualisation and Analysis System Figure 5 6 The Western Isles Analysis Region HIH DO UNT AmE EA Wo Cone w dame Bape TEs The Outputs dialog will then be displayed automatically This allows you to specify the dates and features for which estimates are required see Section 6 4 1 for details Check the tick box next to 40s 80s under Net Change and ensure that both Areal and Linear are checked under Features as shown in Figure 5 7 Estimates and confidence intervals will then be calculated comparing the 1940s and 1980s for both areal and linear features Figure 5 7 Outputs dialog Stock e 1940 lAs e 1980s Net Chang ss T Ue els P40s 80s Feature e Aral wv Linear Interchang M 40s 70s 7 s B0s 40s B s M 95 Confidence Intervals Standard Error OR Cancel The Outputs dialog also allows you to define groups of features which are then combined for analysis Cli
50. timates of net change are calculated assuming a normal distribution Y 1 96 4 var Y 44 National Countryside Monitoring Scheme Visualisation and Analysis System Appendix 5 Square and strata codes Each NCMS sample square has a unique code a number followed by either two or three letters The letters refer to the NCMS district and or region containing the square These are identified in the following table In most cases the same codes are used in the NCMS stratification The exception is Dumfries amp Galloway dag where the stratification does not distinguish between districts NCMS Region NCMS District Code Borders bor Central Clackmanan amp Falkirk ccf Stirling cst Dumfries amp Galloway Annandale amp Eskdale dga Nithsdale dgn Stewartry dgs Wigtown dgw Fife fif Grampian gra Highland Badenoch and Strathspey hbs Caithness and Sutherland cas Inverness and Nairn hin Lochaber hlo Ross and Cromarty hrc Skye and Lochalsh hsl Lothian lot Orkney ork Shetland she Strathclyde Ayrshire and Arran saa Clyde Valley SCV Mid Strathclyde sms North Argyll sna South Argyll ssa Strathclyde Islands SiS Tayside Angus and Dundee tad Perth and Kinross tpk Western Isles ws one square in Edinburgh is labelled led 45 National Countryside Monitoring Scheme Visualisation and Analysis System Appendix 6 Feature codes The underlying NCMS dataset uses short letter codes to identify land cover features rather than th
51. tside squares and whether caution is needed in interpreting the results The figure below summarises the situations when each of the three methods is possible However these are minimum requirements and fairly imprecise results may be obtained if the number of sample squares used is small 40 National Countryside Monitoring Scheme Visualisation and Analysis System Possible methods of analysis For stratum 1 Stratum Method A B or C possible Sample square Analysis Region of interest Method B or C possible but not Method A Only Method C possible mm m I 41 Estimates and standard errors can be calculated from squares falling entirely within the region Unable to estimate the standard error for stratum from the single square falling within the region Unable to estimate the land cover or standard error for stratum 1 without using squares outwith the region National Countryside Monitoring Scheme Visualisation and Analysis System Appendix 4 Calculation of results Terminology Term Explanation S the j NCMS stratum l the intersection between the region of interest and the j stratum A the area of aj the total interpreted area in the k sample square in Y the total area or length of the feature within Vik the interpreted area or length of the feature in the k sample square in lj nj the number of sample squares Suffix Explanation j NCMS strata j
52. ults in the overall standard error for any Analysis Region containing these strata being underestimated In addition confidence intervals will be slightly too narrow Estimates for such Regions will still be unbiased and standard errors will only be substantially affected if these strata comprise a substantial part of the Region This can be assessed using the diagnostic information provided with each analysis If this is the case keep in mind that the true confidence interval is wider than that shown The extent of these strata and their characteristics are summarised below 37 National Countryside Monitoring Scheme Visualisation and Analysis System Extent of strata with only one sample square Characteristics of strata with one sample square Strata NCMS district Type Total area km ork2 Orkney Moorland 147 cas3 Caithness and Mixed but heather dominated 77 Sutherland Note that two other strata also have only one sample square cst1 and tpk4 but in these cases the sample square comprises the full extent of the stratum strata with no sample squares In addition a number of small strata in North Argyll and the Strathclyde Islands were not sampled The location of these and their characteristics are shown below As estimates cannot be calculated for these strata the total area for which results are presented will be slightly less than that of the whole Analysis Region These strata are unlikely to comprise a substantial part of m
53. ysis dialogs allow selection of desired results and estimation method Opens a previously saved set of analysis options and associated Analysis Region Displays on line help The NCMS Square Viewer dialog is shown in Figure 6 1 Select squares from the drop down list in the dialog or by clicking on the square in the NCMS Main View In case of difficulty in selecting squares ensure that the NCMS squares theme Is active slightly raised in the list of themes and the select feature tool is active this is the default when this dialog is opened National Countryside Monitoring Scheme Visualisation and Analysis System Figure 6 1 The NCMS Square Viewer Dialog 2 NCHS Square Viewer Choose square fram list or select square fram NOMS View using the Select Feature tool Selected MCMS Square 249qra Land Cover Chang e 15940 sys Aeris sen Ww 19803 e 403 803 Displa Features M Areal wv Linear Format M Layout Calculate Areas Files ta be created 249gra View 49 gra Layout Layout Template Ad OK Cancel Display options for the selected square are grouped under a number of headings Land Cover Select one two or three dates of interest Change Select one time period to show the major land cover changes between two dates The land cover for these dates will also be selected automatically To undo your selection simply click on any of the checked boxes Display Features Se
Download Pdf Manuals
Related Search