Results and Discussion

Spatial locations of dump site:

A total of 162 grids were covered in the survey carried out during the first year and 696 unauthorized dumps were found to be located in 122 grids.  These dumps were found in all the four quadrants of the city (namely, North East (NE), South East (SE), South West (SW) and North West (NW) inspite of significant differences in the level of development and access to the city centre (Table 1).  The average intensity of dumps was 0.55, 0.69, 0.82 and 1.23 dumps/km2 in NE, SE, SW and NW, respectively.  There are seven grids situated in SW and NW of the city have occurrence of dump location more than 14.  Most of these sites are located at distance of more than 1.5 km from the city’s main road network (National Highways, State Highways and Major District Roads) which connect one end to other end of the city or passes from centre to periphery of the city (Table 2). Out of these abovementioned seven grids, four grids falling inside core city area are more urbanized with high proportion of built up area compared to other three grids outside the core city boundary where the grid area is predominantly represented by open land. 
Contrary to expectations, 303 dumps were found within core area of the city (Table 1). 

The average intensity of the spread of such dumps was 2 dumps/km2.  These dumps were distributed non-uniformly in the four quadrants of the city (Figure 2a).  The number of dump sites determined were 27 (NE), 10 (SE), 54 (SW) and 212 (NW).  The number of dump sites was highest in NW although the NW quadrant is the closest quadrant to the only authorized waste processing site in Bangalore. 

Bangalore city administrative jurisdiction was redefined in the year 2006 by merging 8 neighbouring Urban Local Bodies (ULBs) and 111 Villages of Bangalore Urban Districts.  Many of these areas have not been urbanized and the character of land use and intensity of use in many parts of the outer area closely resembles the area in the periphery of Bangalore city even before their being included into the city.  Therefore these two zones have been taken up as one in this study to interpret the observed results and needs to be used for other cities too.  The unauthorized dumps located outside core city area were 393, distributed in all four quadrants of the city (Table 1).  The average intensity of the spread of the dumps per grid was 4 (1 dump/km2).  Most of the dumps are located near to the main road network which can bear the weight of waste loaded trucks and also close to vegetation covered areas, hilly regions, barren unused land or water body.  The number of dump sites were 90, 143, 125 and 35 in NE, SE, SW and NW quadrants, respectively.  SE and SW quadrants of Bangalore city do not have any centralized USW processing and disposal facilities nearby.  Whereas, the NE and NW are closer to existing centralized waste transfer and waste disposal sites such as Mandur and Mavallipura, respectively (Figure 2b)

Therefore the higher numbers of the dumps in the SE and SW quandrants of Bangalore may be attributed to the greater distance of transport to the processing site and resulting reluctance by transporters or related system to carry it away to the logical place.  Transporting a truck load of USW through the city to the permitted disposal site, Mavallipura, could take as much as two hours while these sites are only 20-40 minutes away in the northern quadrants.  The permitted processing facility closes at sunset.  Collection vehicle loaded with waste delayed beyond 4.30 – 5.30pm end up parking the fully loaded truck overnight at the gates of the processing site.  This potentially becomes one of the major reasons for waste transporter to seek these unauthorized dump sites. 

Composition of wastes in dump sites:

Out of the possible six types of dominant wastes found at dump sites described under methods, 73.9% of the total number of locations had construction debris and plastic wastes and to a smaller extent organic wastes.  The predominant component of wastes recorded in unauthorized dump sites was construction debris followed by plastic and organic waste.  Considering that a large part of the unauthorized dumps contain construction debris, it is imperative that alternative uses for them needs to be evolved to recycle them back into the construction industry.  This is important because India and more importantly Bangalore faces a significant shortage of construction materials mainly sand and aggregates and therefore recycling construction wastes back into the construction industry can make this industry a lot more sustainable [35].  Plastics and organic wastes occur next in the order of dominance among 25.3% and 23.3% of total dump locations, respectively.  Bangalore city houses a very large plastic recycling industry dominated by small units (1 tpd scale) and such a spread of plastic wastes therefore suggests that it would be more sustainable to locate these small units where plastic wastes dominate.
The dump sites located inside core area had organic wastes followed by construction debris and plastic (Figure. 3a), respectively, and outside core area major contribution came from plastic and construction debris (Figure. 3b).

From the dominant waste type found at the dump site and possible reasons for their occurrence it may be surmised that within the core area of Bangalore city, these dump sites occur due to enefficient and improper collection as maximum number of dump sites predominantly have organic wastes and the composition closely resembles fresh domestic wastes [16].  The presence of food and kitchen wastes, highly perishable wastes, shows that these dumps are ephemeral and are cleared within a weeks time.  The organic fraction dominant unauthorized dump sites outside the city have a somewhat different character.  The presence of organic wastes components at various stages of decomposition suggested that these dumps are not ephemeral in character.  These are long lasting dump sites in comparison to dumps situated inside the core area.  After a certain point of time, either due to large size of dump site or due to the loss of cover and anonymity they are abandoned. 

Size wise distribution of dump sites:

Unlike developed countries, Indian waste (especially freshly collected wastes in Bangalore) has a large fraction of decomposable organic matter [16].  Most of organic matter decomposes while leaving behind predominantly the non-decomposable fraction that gradually spreads over the area and becomes unaesthetic.  Table 3 shows the mean, range and standard deviation of physically measured area estimated at 33 locations for construction and non construction waste categories.  The greatest area was recorded for construction waste category.  The land area occupied by these dump sites varies significantly in size.  Table 4 present the mean, range and standard deviation with three size classes chosen for waste spread area namely, <200, 200-400 and >400 m2.  Waste dumps which are in class of >400 m2 occupy greatest area.  This suggests that there is pattern in the dumping at each of these locations and these locations also remain functional over a specific period of time.  These dump sites are used at irregular hours and the act of unauthorized dumping cannot be easily detected and monitored.  Therefore regular survey or assessment of dump sites area can assist in making decisions on planning and improving the quality of solid waste management by the city authorities. 
A physical survey that involves measuring the length and breadth of all of such dumps is not possible because, as mentioned under methods, people residing nearby these ‘unauthorized’ dumps often vehemently oppose such surveys and volunteers can suffer harm.  Second, as material dumped is loose, physical measurements can lead to injury and even death.  Therefore, alternatives need to be developed.  In order to overcome these survey difficulties, in this study the spread area of 33 dump sites were assessed by three methods of SM, VE and GE based assessment.  Using coefficient of variability (CV) as an index [19, 23], the VE method of area estimation was found to be the most accurate having a CV of 2%.  A higher CV value (5%) has been found for GE method.  The correlation coefficient in Table 5 for expert’s estimate with VE is more than that with GE.  Both these measures show weak correlation with the physically step measured area.  In both types of area estimation, there is a decrease in correlation coefficient with an increase in the dump size except for 200-400 m2 category of waste dump sites.  This suggests that smaller dumpsites are estimated more accurately compared to the large dump sites.  The correlation coefficient for VE is -0.538 and 0.300 for dump sites of <200 m2 and >400 m2, respectively.  Similarly for GE method, the correlation coefficient is -0.386 for unauthorized dumps of size <200 m2 whereas for unauthorized dumps of >400 m2, it is -0.217.  In general, there are both under estimation as well as over estimation in area for the different categories of waste.  A similar phenomenon where agriculture related observers have over-estimated low herbage volume and under estimated cases with high herbage volumes [36].  From this study, it appears that a higher accuracy of assessment can be achieved in two ways.  Firstly ensuring that volunteers are trained better and exposed to a larger sample size prior to visiting actual dump sites.  On the other hand, this approach could still be used however, the area needs to be classified into at least three classes as done in this study with varied accuracies.  Therefore, it is concluded that VE can be used reasonably well for smaller areas.  Visual assessment of dump area is an efficient, rapid and non-risky method for determining the area of USW spread at these unauthorized dumps.  However, continuous calibration of the observer’s estimates is required to maintain the method accuracy.

Temporal change in dump sites

New dumpsites are being created constantly in this study area while older dumps are being abandoned after a certain period of usefulness.  Dump sites inside core area of the city are ephemeral and their contribution to the overall waste-spread area accounts for <20%.  Therefore the second survey that examined the extent of earlier dump sites used in subsequent year was restricted to outside core area of the city.  In the second year survey, a total of 452 locations were visited in the very same grids which were visited earlier by student volunteers of the first year.  It includes 125 new dump locations and 327 old dump locations (Figure 4). 

The site where 128 earlier dump site existed had now become locations for the construction of new buildings.  The average intensity of dumpsites were 0.69, 1.1, 0.69 and 0.41 dumps/km2 in NE, SE, SW and NW, respectively.  Compared to the first survey, the frequency of dump site per grid in the second survey had reduced in SW and NW quandrants, whereas, there was only a small reduction of dump sites in NE and SE quadrants.  A paired t-test was performed to check grid-wise change in number of dump locations for the two consecutive years of field survey.  This confirms that there is a significant change in grid-wise locations observed in two complete survey conducted outside core city area at p<0.05 (α = 0.05, n = 87).  Thus there is a continuous temporal change in waste dump locations and we need to adopt methods which can locate dump locations in minimum time and possibly the lowest cost. 
The results of this study demonstrated that such a method was preferable to identify locations of unauthorized dumping as well as quantify the compostion and extent of wastes spread thorugh dump sites.  However, there is a lower potential to accurately determine spread of the dump sites through visual methods.  Thus, with this approach an assessment can be replicated and applied to other cities.