Characterization of short-term salinity fluctuations in the Neretva River Delta situated in the southern Adriatic Croatia using Landsat-5 TM

摘要

Elevated soil and water salinity can lead to increased risk of crop yield and quality reduction as well as land degradation. This is especially emphasised in arid and semi-arid regions with high levels of evaporation and irrigation practice. In this study, we assessed the potential of applying Landsat-5 Thematic Mapper (TM) to evaluate low to moderately salt-affected agricultural areas of the Neretva River Delta, situated in the southern Adriatic region of Croatia. Monthly measurements of water electrical conductivity (ECw), sampled from drainage/irrigation canals, were used as in situ salinity indicators. The relationship between agricultural plots reflectance (32 Landsat-5 TM images, distributed over 22 months from May 2009 to October 2011) and water salinity (measured at ten monitoring stations) was analysed using six Landsat-5 TM bands and spectral indices. Correlation analysis, simple (SLR) and multiple linear regression (MLR) models were implemented to predict ECw values using the satellite data. Very strong correlations were achieved in the spatial domain for August 2009, September 2010 and September and October 2011, while correlations in the temporal domain had relatively lower values. The Landsat-5 TM data managed to explain 43% of ECw variance using the SLR model, and 62% of the variance using the MLR model with three regressors. Landsat-5 TM remotely sensed data, paired with the in situ observations, showed a promising potential in enhancing the quality of environmental monitoring of low to moderately salt-affected agricultural fields in the irrigated deltas.