The Use Of The Buffer Function to Study And Compare The Prevalence Of Long Term Illness Aling The Motorways In West Yorkshire Region In The United Kingdom Mr Ankur Das System Analyst & GIS Officer Trans Virtual Pvt Ltd 1st Floor, Institution of Engineers Bldg Pan Bazar Guwahati- 781008 Tel No: +91-361-2604989 /2608805/ 2730664 Mobile: 9854049156 ankurdineshdas@yahoo.com ankur@transv.net Abstract: This work builds upon preliminary studies which showed a high count of long term illness among the population placed along the nation’s motorways. Through this work, an attempt has been made to quantify and map the prevalence of long term illness along a 5 Km transect on either side of the motorways for the West Yorkshire region. Results achieved have been used to come to conclusions as to whether they tally with earlier reports or make a different picture INTRODUCTION In this study that follows, the aim has been to see the existing trend of long-term illness within and outside of a 5 km buffer zone. This buffer zone has been measured and determined with respect to the motorways that are present within the region. Datasets for the purpose of studying long term illness having relevant variables have been obtained (downloaded) from the CASWEB website, so also spatial data relating to Motorways have also been obtained from the same website. It has been attempted to quantify the number of people who have developed long-term illness in a buffer zone 5 km on either side of the motorways and compare it with those outside the buffer to see if more people are ill near the roads. It is presupposed from the results of earlier studies that there is a high prevalence of long term illness in the vicinity of the motorways. There have been many debates on the possible factors that might be responsible for this spatial condition. This study however instead of delving into the factors responsible has been kept within the stated aims to map the long term illness within the vicinity of the motorways and test the felicity, the results of earlier studies whether the prevalence of long term illness is more prominent in the vicinity of the motorways as compared to the other regions within West Yorkshire. One important pre-condition that has been accepted and followed in this study is the removal of all people aged 65 and over from the datasets. This is because it is seen that most of the people who fall in this category usually live in retirement homes, which are normally located near the motorways. This category of people generally has a high count of long-term illness as they are aged and hence affects the analysis. So, in order to make the study more in tune with existing economic activity, it has been decided to incorporate all classes of the population who are under the age of 65. Henceforth, the results that have been achieved and documented in this report should be read as not involving the people aged over 65, expressed as 65+ in the succeeding paragraphs. THE DATA This report concerns the West Yorkshire region. Tables have been selected from the CASWEB website, which have measure of long-term illness as one of their attributes. Tables having population statistics have also been selected. The measure of their attributes has been selected at the ward level of West Yorkshire and downloaded in the mappable Arcview shapefile format. The tables and the selected attributes are as follows- From the CASWEB Database,
The statistical analysis and allied operations have been done on ArcGIS and ArcView software alternately and the resulting maps generated. DATA EDITING The population over the age of 65 have been removed before any analysis has been done. This involved addition of new fields in the main attribute table due to the statistical operations performed. As the data was in the Arcview shapefile format, ArcMap did not allow editing, so the changes and calculations were done in ArcView. Before this was done, it was attempted to create a coverage in ArcCatalog by converting the shapefile. But this did generate the coverage but there was loss of data in the attribute table. Most of the attributes did not have any other measure other than 0. So, all the modifications were done in ArcView and then the changed dataset was opened in ArcCatalog and main component of the spatial analysis done. Calculations have been performed on the attributes of the three tables mentioned earlier and in the process new fields have been created in the main attribute table for the entire County. Firstly, the total population for each of the wards has been calculated. This has been done by adding up the fields which have population greater than 65, and subtracting the resultant value from the field denoting the total population of the wards for all age groups. So, we get the total population in each ward minus the over 65 years population count. The steps can be represented as Age-65 = (Allages) - (Tot65+) where, ‘Allages’ is the total of all age groups for the wards ‘Tot65+’ is the sum of the population count which are above the age of 65 ‘Age-65’ is the resultant total population count for the wards minus the 65+ people. Secondly, the total long-term ill people for the wards have been calculated which is without the 65+ count. For this, the variables in the tables denoting the long-term ill in households and communal establishments have been considered. Here the total long term ill in each ward have been calculated by subtracting the long-term ill which are above the 65+ category. Similarly, the total long term ill in communal establishments have been calculated by subtracting the long-term ill above the 65+ category. Then I have gone on to calculate the total long-term ill for the wards by summing the above two values. The calculations can be represented as follows- ill-65H = AllagesH – {(65-74H) + (75+ H)} where, ‘AllagesH’ is the total long term ill in households for all the age categories in the wards. ‘65-74H’ represents the long term ill in households who are aged between 65-74. ‘75+ H’ represents the long-term ill in households who are above the age of 75. ‘ill-65H’ represents the total long-term ill in households minus the 65+ population. Similarly, ill-65C = Allages – {(65-74C) + (75+ C)} where, ‘AllagesC’ is the total long-term ill in communal establishments for all age categories in the wards. ‘65-74C’ represents the long-term illness in communal establishments who are aged between 65-74. ‘75+ C’ represents the long term ill in communal establishments who are above the age of 75. ‘ill-65C’ represents the long-term ill in communal establishments minus the 65+ population. The total long-term ill ‘Totalill’ for the wards has then been calculated by adding the fields of ‘Ill-65H’ and ‘Ill-65C’ which is represented by , Totalill = (ill-65H) + (ill-65C) After having derived the wanted count of total population and that of long term ill, the analysis has then been performed in ArcGIS. ANALYSIS & RESULTS The ‘motorways.zip’ contains the motorways for the whole country. Initially with the help of Arctoolbox, the portion of the motorways that lie in the West Yorkshire County have been clipped and a coverage generated. The shapefile of the West Yorkshire region is then opened in ArcCatalog. Its attribute table contains all the new fields created in ArcView with their respective counts. The shape files are opened in ArcMap and the analysis done. Three maps have been generated Map 1, Map 2 and Map 3. In Map 1, the total population and the total long term ill for each of the wards for the entire West Yorkshire County has been represented. The figures have been shown as measures in ranges. This is a very simple map and shows the basic data representation. For generating Map 2, the motorway coverage has been overlaid on the West Yorkshire shapefile and the 5 km buffer generated and then the long-term ill figures were generated as ranges. Map 2 hence, shows the 5 km buffer zone, the motorways contained, and the long term ill for both inside and outside the buffer zone. This map answers the question that; yes, there is a definite positive difference between the number of long-term ill persons in proportion to the whole population for the regions inside and outside the buffer. The map definitively shows that the number of long term ill is more within the buffer zone than outside. This is in agreement to the results achieved by earlier studies that people living near the motorways are more prone to long-term illness. It will be worthwhile to note here that majority of the people of the entire West Yorkshire region live within this buffer zone. Map 3 represents only the buffer zone along with the long-term ill within it as a measure in ranges. The attribute table of the clipped coverage used in generating this map gives the statistics of the population and long-term ill within the buffered zone. The population and long-term ill statistics for both inside and outside the buffer zone are as follows-
For the area within the buffer, the absolute value of long term ill is 1,23,055. So, the absolute number of normal people within it will be the total population minus the long term ill i.e. 15,33,272 – 1,23,055 = 14,10,217 Similarly, for the area outside the buffer zone, as the absolute value of long-term ill is 13,360, the absolute number of normal people will be the total population minus the long-term ill, i.e. 1,69,371 - 13,360 = 1,56,011 Discussion & Conclusion It is not only graphically seen but the statistics also point that there is a large positive difference between the number of long term ill persons inside the 5 km buffer with those outside it. These results definitely are in accordance to earlier results that more people are ill near the motorways. In order to make the study more reliable the 65+ aged population had been removed in the beginning from the dataset as these groups primarily live in retirement homes, which are usually placed near the motorways. But even after this category has been removed from the dataset we still get results, which directly points to the fact that more people are prone to long-term illness near the motorways. An important point to note is that almost 90.1% of the total population of West Yorkshire wards are concentrated in this buffer zone of 5 kms. Also 90.2% of the total long-term ill persons for the County are also found to be concentrated in this region. As the major portions of the population are concentrated in the buffer zone, it is also likely that characteristics of the population such as long-term illness will be also highly reflected. The presence of more population within the buffer zone also points that more household as well as commercial and communal establishments are in it. And hence the high measure of long-term illness. It is quite logical to set up residences and business activities in places, which are near the motorways as location is an important determinant of any economic activity. This certainly denotes that the urban and working environment for the Counties’ population are also within this zone. Hence for such an urban environment, other demographic factors also come into play as also social and economic factors such as deprivation and poverty. It has been a focus of many studies to understand the relationship between deprivation and long-term illness. And most studies have been found to suggest that people living in deprivation are more prone to report long- term illness than others. But these are only suggestive and not the accurate factors for the results, as there can be other important ones as well. If the 65+ age category were not left out, the results would have been no different. It would have only stressed on the current results more positively. However the results obtained from this report should be treated only as a single study and should not be taken as an indicator for the situation prevalent for the whole country. Similar analysis of other Counties by the same methodology may yield very different results from the ones documented in this report. Apart from the methodological problem regarding the management of the main attribute table of the shapefile mentioned earlier in this study, there is one more that deserves pointing out. The population statistics for the buffered zone is calculated from the coverage that is a clipped output from the coverage representing the whole West Yorkshire wards. When this clipping is done by the means of the geoprocessing tool in ArcMap, the buffer zone is obtained as a separate coverage as desired. And each ward that falls in the buffer zone takes the attribute value that is relevant for the whole of the ward. This is in conflict because it is seen that the buffer cuts through half and sometimes very small portions of the wards, but these wards instead of taking the measures of the attributes for the concerned portion of the wards take the value, which represents the entire ward. This certainly has an effect on the result, as accurate attribute measures certainly would have made a difference on the result. But this problem could not be avoided as the initial shapefile downloaded from CASWEB was at the ward level. If it were in a finer level this problem would have been alleviated to a great degree. Then, the analysis would have taken an entirely different direction. But however this issue could not be taken into consideration and the results have been obtained nevertheless. In conclusion, this report has tried to make a comparative study of the levels of long term illness within and outside a buffer zone based on datasets downloaded as well as reference information provided. The results obtained in the process have pointed to the existing knowledge that more people are ill inside the buffer zone. This strengthens the existing arguments that are responsible for this condition and also pinpoints to certain other relevant factors that have been mentioned in the study. ILLUSTRATIONS: Map1   
|