Algae bloom detection and its spatial temporal change quantification in Lake Victoria.
Abstract
Lake Victoria is an important geographic feature being the backbone of various economic activities such as fishing and agriculture. It is a source of water for consumption for very many people and supports the ecosystem in its region. Several difficulties exist in and around the lake environment, based on different causative factors. Inputs of nutrients into the lake, for example, has led to eutrophication and algae development. Lake Victoria's water quality has degraded drastically over the last few decades as a result of algal concentration in the water that has grown to noticeable quantities. (Njiru et al., 2008). For effective monitoring of massive algal blooms, a 2 weekly examination routine is recommended by the World Health Organisation. In situ-based sampling and laboratory tests on the other hand are expensive and time consuming and hence a better strategy that’s financially feasible and a good data collection method ought to be used. The few in situ observations of phytoplankton biomass make it difficult to also characterize the temporal and spatial variability of the cyanobacteria (algae) blooms. (Visitacion et al., 2019). This research therefore sought to explore the use of satellite imagery in monitoring algal blooms.
The floating algae index is an ocean colour index that was engineered to detect floating algae in open ocean environments. ‘The FAI, defined as the difference between Rayleigh-corrected reflectance in the NIR and a baseline formed by the red and SWIR bands’ (Hu, 2009). It was utilized of the 10 area scenes to generate algal distribution layers and were then validated. In comparison to conventional methods where ocean colour is validated with in situ measurement, and sea surface temperature, it is complex to directly validate algal blooms from satellite data using field measurements. This is because of algae being patchy and it is complex to obtain near timely field recordings of the algal distribution. Accuracy assessment was carried out through comparison of the trends of algae processed in SNAP and visual inspection with high-resolution sentinel 2 sensor true colour and false colour imagery that had a spatial resolution of 10m (Visitacion et al., 2019).
A total area of 2039.12km² of Lake Victoria north western shelf was evaluated in this algal spatial distribution study and constituted of 1sentinel image scene clipped to the lake area. The largest algal coverage was seen to constitute to 10.75% of the lake area under study. From sample points the highest algal concentration was extracted to be 0.0165 but from all general pixel statistics, it went as high as 0.5 as the FAI reading indicating thick algal scum. The largest algal change occurred between February to March. There was a temperature change of 0.7°C/k and a percentage algal change of 7.79% which further showed that algal growth is temperature dependent. However, when the correlation was evaluated, it showed a very low positive correlation i.e. R² = 0.0334. Several hotspots were obtained when hotspot analysis was conducted and constituted of Bunjako bay, Tonde bay, Nakiwogo, Gobero bay and port bell all flagged at a 95 - 99% confidence level as portrayed in Figure 4. This study therefore portrays and gives water authorities valuable information to conduct research on causative factors leading to rise in algal growth most especially next to Land masses.