A limnological study of Lake Patzcuaro, Mexico with a consideration of the applicability of remote sensing techniques

Abstract

This thesis presents the results of a hydrobiological study with implications for lake management obtained during six months of field work in Lake Patzcuaro, Michoacan, Mexico. The conventional data are coupled with surveys made with multispectral satellite imagery. Lake Patzcuaro, Michoacan, Mexico is a tropical high altitude freshwater lake located in the southern edge of the Mexican plateau. The lake originated from tecto-volcanic processes associated with the Eje Neovolcanico Transversal during the late Tertiary and Quaternary periods. Data for 25 morphometrical parameters describing the geomorphology of Lake Patzcuaro were derived from an up-to- date bathymetric survey with contour lines at 1.0 m intervals: Maximum depths were observed in the northern part of the lake whereas considerable shallow areas are developing in the south: insulosity is continuously modified due to certain areas becoming part of the mainland and the appearance of new islands with increasing shallowness and man-made channels. The lake is an endorheic basin and the annual water balance is primarily controlled by differences between rainfall and evaporation. Since annual evaporation is higher than rainfall, it is considered that net water inputs to the lake come as seepage from the drainage area. However, by using a simple mass balance simulation model it is illustrated that the contribution of seepage is being reduced due to the continuous deterioration of the catchment area. The lake is well mixed and high oxygen levels are always present from surface to bottom. The lake is alkaline- carbonate with moderate levels of hardness and high concentrations of total phosphorus, chlorophyll-a and suspended solids. By comparing three conventional trophic state indices and input-output phosphorus models, Lake Patzcuaro is considered to be mostly eutrophic with some isolated mesotrophic and, oligo-mesotrophic areas. High turbidities are observed possibly due to volcanic silt, raw sewage and increasing erosion loads. It is considered that Lake Patzcuaro is principally 1ight-1imited due to scattering and attenuation effects from inorganic materials. A predictive model of water quality variables has been developed for SPOT-1 multispectral imagery, and applied.to Lake Patzcuaro. By using canonical and principal component analysis it is shown that, at most, two water quality variables, chlorophyll-a and suspended solids concentration, can be derived from the SPOT data. Having established the independent predictiviness of a set of empirical relations between the SPOT data and the water quality parameters, the whole lake was analysed to reveal the spatial distribution of chlorophyll-a and suspended solids. It is revealed that a very dense algal bloom had occured during the period of study. No bloom of such intensity has ever been observed before in Lake Patzcuaro. The applicability of SPOT imagery to water quality monitoring is clearly demonstrated. The cost-effectiveness of remote sensing techniques for water quality mapping and lake management is discussed using the Lake Patzcuaro as an example.