Development of a near surface remote sensing system for water quality parameter estimation
In the MuDak-WRM project a near surface remote sensing system is developed to support the qualitative and quantitative measurement of material input to water reservoirs. This subproject addresses the scale leaps between satellite remote sensing with reservoir wide coverage and localized distributed in-situ measurements. The latter enables the exact determination of all parameters needed, but sampling is only feasible at single points, time-consuming and cost-intensive. Using multi spectral satellite data enables an automatic, extensive analysis, but usually with less accuracy. In comparison to other methods satellite images suffer from low spectral resolution and an atmospheric correction is inevitable. In addition the data acquisition is not flexible and depends on the cloud coverage.
In this project a near surface remote sensing system composed of a hyper-spectral and a thermal infra-red imaging sensor on a UAV is employed to observe the material input at the inflow of the reservoir and the proliferation in the reservoir. Especially the episodic input at the shore will be investigated, in particular after heavy rainfall. The benefits of using a UAV are the low costs and the flexible acquisition of data providing high resolution in space, time and spectrum. Optically active parameters like the amount of total suspended solids (TSS), coloured dissolved organic matter (CDOM) and chlorophyll-a are the primary parameters of interest. Beyond that, existing methods are used and adapted to infer optical inactive parameters like total phosphorus concentration (TP).
In view of an operational and commercial implementation of the developed methods, the use of autonomous UAV based remote sensing is expected to support the time and cost intensive in-situ sampling. The objective is to develop a concept of an optimised minimum monitoring system to complement in-situ sampling with hyper spectral near surface remote sensing and multi-spectral satellite data.
Two drinking water reservoirs and their catchments will be investigated in this research project, the reservoir Große Dhünntalsperre in North Rhine-Westphalia (Germany) and the Passaúna reservoir in the federal state Paraná of Brazil. Comparing the deliverables guarantees the transferability of insights to further reservoirs.