Estudio de la clorofila-a satelital en el golfo San Jorge: validación con datos de campo y comparación intersatelital

Abstract

En la presente tesis se estudió la estimación de Cl-a de origen satelital generada por distintos sensores y se la comparó con mediciones in situ en el Golfo San Jorge (Argentina). Se evaluaron datos de MODIS/Aqua y VIIRS/Suomi-NPP para los años 2014 y 2016, mostrando correspondencia aceptable (r2 = 0.59-0.82). Además, se compararon estimaciones de Landsat 8 y Sentinel-2 con MODIS/Aqua para los años 2017-2019, encontrando variabilidad en las correlaciones. Se destaca la necesidad de más datos empíricos para ajustar algoritmos regionales. Como aporte concreto a la misión SABIA-Mar, tanto los protocolos de comparación como los estadísticos para establecer medidas de errores, serán incluidos en los planes de calibración y validación de los productos de la misión.

In this thesis, Chl-a derived from different sensors was studied and compared with in situ measurements in San Jorge Gulf (Argentina). Methodologies were developed and different statistics were calculated to compare these data. Initially, Chl-a in situ measurements obtained during campaigns in 2014 and 2016 were compared to Chl-a estimated by the OC3 algorithm using MODIS/Aqua and VIIRS/Suomi-NPP. The results showed an acceptable correspondence between field measurements from both campains with the values estimated by MODIS/Aqua (r2= 0.59-0.67 respectively) and VIIRS/Suomi-NPP (r2= 0.54-0.82 respectively). On the other hand, Chl-a derived from Landsat 8 and Sentinel-2, was compared with Chl-a from MODIS/Aqua (used as reference for satellite/in situ comparisons) in coinciding scenes during 2017, 2018 and 2019. Correlations varied, with some cases showing good correspondences and others weak. Based on the results obtained in this thesis, it is evident the requirement for more empirical data to develop or modify regional algorithms. As a specific contribution to SABIA-Mar mission, both the comparison protocols and the statistics to determine error measurements, will be included in the calibration and validation plans for the mission products.