Lop Nur, a playa lake located on the eastern margin of Tarim Basin in northwestern China, is famous for the "Ear" feature of its salt crust, which appears in remote-sensing images. In this study, partial least squares (PLS) regression was used to estimated Lop Nur playa salt-crust properties, including total salt, Ca , and Fe 2+ using laboratory hyperspectral data. PLS results for laboratory-measured spectra were compared with those for resampled laboratory spectra with the same spectral resolution as Hyperion using the coefficient of determination (R 2 ) and the ratio of standard deviation of sample chemical concentration to root mean squared error (RPD). Based on R 2 and RPD, the results suggest that PLS can predict Ca 2+ using Hyperion reflectance spectra. The Ca 2+ distribution was compared to the "Ear area" shown in a Landsat Thematic Mapper (TM) 5 image. The mean value of reflectance from visible bands for a 14 km transversal profile to the "Ear area" rings was extracted with the TM 5 image. The reflectance was used to build a correlation with Ca 2+ content estimated with PLS using Hyperion. Results show that the correlation between Ca 2+ content and reflectance is in accordance with the evolution of the salt lake. Ca 2+ content variation was consistent with salt deposition. Some areas show a negative correlation between Ca 2+ content and reflectance, indicating that there could have been a small-scale temporary runoff event under an arid environmental background. Further work is needed to determine whether these areas of small-scale runoff are due to natural (climate events) or human factors (upstream channel changes).