View Abstract


There are 3820 lakes greater than 1 ha in New Zealand and in-situ monitoring of the water quality of merely a fraction of these represents considerable logistical, financial and data-handling challenges. Remote sensing by satellites provides imagery of the entire country at return periods of days to weeks and algorithms exist to derive water quality parameters from their data streams. Thus, remotely sensed data may contribute efficiently to nationally legislated water quality monitoring requirements, but satisfactory performance of water-quality parameter retrieval algorithms has to be demonstrated first. The main source of errors in water quality parameters derived from remote sensing reflectance is the interference of optical signatures from constituents such as chlorophyll a, suspended sediments and colored dissolved organic matter. New Zealand’s versatile geology, climatic gradients and range of anthropogenic pressures produce a wide range of optical lake water types; in combination with a scarcity of in-situ observations, it is currently impossible to generalize about the quality of remote sensing retrieval algorithms. We present a New Zealand-wide assessment of the variability of lake water colour using chromaticity coordinates from the entire archive of Landsat 5 and 8 reflectance and relate this variability to land-use characteristics of lake catchments. This provides the foundation for developing a suite of algorithms for the retrieval of lake water quality parameters across New Zealand’s optically diverse lakes and helps regional stakeholders to determine whether remote sensing presents a worthwhile opportunity for the routine monitoring of lake water quality.


Lehmann, M. K., University of Waikato, New Zealand,

Allan, M., University of Waikato, New Zealand,

Nguyen, U., New Zealand,

Hamilton, D. P., University of Waikato, New Zealand,


Oral presentation

Session #:031
Date: 02/28/2017
Time: 14:45
Location: 323 A

Presentation is given by student: No