We know, humans are not really good at caring or even individually perceiving their impact on the surrounding environment. As such we are doing very good at polluting rivers around the world. Some rivers have even been known to be flammable at points. One of the metrics to measure pollution levels and pin-point major pollution sites is total suspended solids or TSS for short. This is basically how murky the water is because of the stuff floating in it.
The conventional way of taking this metric is to go around the river and take water samples. This can be effective for smaller rivers, such as the ones we have here in Estonia, where the breadth of the river in the selected region can be effectively sampled in a reasonable amount of time by a limited amount of people. Extrapolating this method to gigantic rivers crossing metropolitan environments, however, means that gathering and analysing a larger amount of samples takes a large amount of workforce and exponentially more time. This leads to spotty data both temporally and spacially, giving the first reason to use remote Earth observation data to try and develop new methods to acquire TSS data.
Then came COVID, and suddenly the workforce gathering the data was no longer available. This was especially the case in China where river pollution is among the highest in the world and a complete lockdown was quickly put into effect. As such, a unique opportunity presented itself to further develop Earth Observation based detection of TSS and investigate both the real magnitude of human activity and the impact of the lockdown on it.
For this, two sections of the Min River crossing through the Fuzhou urban area were taken as research specimens. A previously developed algorithm for TSS estimation in the Min River, the “Wen algorithm”, which used Landsat 5 data, was taken as a basis and compared to two more generic algorithms, Nechad and Novoa, which were found to be comparable in results. The Wen algorithm was then calibrated to compensate for atmospheric changes over time and applied to contemporary data from Landsat-5 and 8 and compared once again to Nechad and Novoa, and found to perform well.
As for the change in TSS due to lowered human activity, they also overlaid the data onto the images of the river, where it is severely visible.
For a conclusion it was proven that Earth Observation data could be used to reliably to replace other more conventional methods for measuring TSS and it was very effectively shown how much impact industrial operations from very specific city regions have on water quality very far downstream.