Objective 4: To identify strengths and limitations of the study approach.
The study has several fundamental strengths and weaknesses. First, our data was sourced from reliable organizations such as the Ministry of Natural Resources and Forestry. The data we will be using were created very recently, and the selected scale is appropriate for the purposes of our study. Some other strengths of our study include easy access and low costs for data collection, as well as suitable sample size with replicates.
Conversely, a few limitations emerged. Water samples reflecting water chemistry and the LANDSAT imagery were only taken on two days during the 2018 summer. Trends and relationships between seasonality and features, metal concentrations, vegetation and wetland cover can not be established due to the possibility that water chemistry varies throughout summer. It might be dependent on the climate, water temperature, water conditions (clear, cloudy) and environmental factors (presence of sewage for example). There is also a possibility for minor data quality issues resulting from atmospheric interference that may have disrupted signals and transmissions for LANDSAT imagery. Inaccuracies may arise from these data as cloud cover and weather dictate the quality and accuracy of the resulting data, the magnitude of which is also dependent on the band and its wavelength range. Additionally, the time of year affects the vegetative and wetness index results as the processes behind the expansion and growth of these parameters is highly seasonal. Conducting the survey in July might have missed the maximum vegetative cover and many wetlands dry up during this period. Given these limitations and the timing of other surveys of this location, it is reasonable to proceed under the assumption that this timing accurately depicts the vegetative state and wetland cover of these catchments. Furthermore, the MCE model has a few drawbacks itself. First, standardization of criteria scores shows the degree of suitability: the higher the scores, the more suitable the areas. Nevertheless, there is no actual threshold that determines which areas to be included and which areas to be eliminated (Drobne & Lisec, 2009). Another limitation of the model is that it only considers raster data. We considered catchment area as one of the catchment characteristics initially but since area is vector data, it cannot be analyzed by MCE and had to be excluded from the study.