Scale Effects on Chlorophyll-A Concentration Retrieved: Assessment and Validation Using Indian Remote Sensing Satellite

To understand the scale effects on chlorophyll-a (chl-a) concentration retrieved from satellite images, the two-band algorithm (TA) and three-band algorithm (TBA) were constructed for estimating chl-a from satellite images. Two synchronous images of Advanced Wide-Field Sensor (AWiFS) and Linear Imaging Self-Scanner (LISS) of Indian remote sensing satellite were used to assess and validate the scale errors of these two algorithms. They were collected at local time 02:55:46:471 and 02:58:25:053 on October 8, 2005 in Yellow River Estuary, and their spatial resolutions are 24 m and 56 m, respectively. From the results of this study, it was found that: (1) the relative scale error (RSE) of TA and TBA, caused by scale changing from LISS to AWiFS, varied from 0% to 100%; (2) the RSE was correlated with the spatial non-homogeneous degree of chl-a distribution; and (3) using TBA to estimate chl-a concentration in Yellow River Estuary decreased 2.55% of model uncertainty, but increased 4.97% of scale errors, in comparison with TA. Additionally, the study indicated that the performance of algorithms for chl-a estimation was greatly affected by the scale error. If the scale effects of chl-a retrieval algorithm were taken into consideration, TA had a superior performance to the TBA in this study.