Automatic Parameter Optimization for Support Vector Regression for Land and Sea Surface Temperature Estimation From Remote Sensing Data

2009 ◽  
Vol 47 (3) ◽  
pp. 909-921 ◽  
Author(s):  
G. Moser ◽  
S.B. Serpico
Keyword(s):  
Remote Sensing ◽  
Surface Temperature ◽  
Sea Surface Temperature ◽  
Support Vector Regression ◽  
Parameter Optimization ◽  
Remote Sensing Data ◽  
Sea Surface ◽  
Support Vector ◽  
Temperature Estimation ◽  
Sensing Data

scholarly journals Removing Striping Noise from Cloudy Level 2 Sea Surface Temperature and Ocean Color Datasets

2016 ◽  
Author(s):  
Brahim Boussidi ◽  
Ronan Fablet ◽  
Bertrand Chapron
Keyword(s):  
Remote Sensing ◽  
Wavelet Transform ◽  
Surface Temperature ◽  
Sea Surface Temperature ◽  
Ocean Color ◽  
Remote Sensing Data ◽  
Sea Surface ◽  
Discrete Wavelet ◽  
Sensing Data ◽  
Fourier Filtering

This paper introduces a new destriping algorithm for remote sensing data. The method is based on combined Haar Stationary Wavelet transform and Fourier filtering. State-of-the-Art methods based on the discrete wavelet transform (DWT) may not always be effective and may cause different artifacts. Our contribution is three-fold: i) we propose to use the Undecimated Wavelet transform (UWT) to avoid as much as possible shortcomings of the classical DWT; ii) we combine a spectral filtering and UWT using the simplest possible wavelet, the Haar basis, for a computational efficiency; iii) we handle 2D fields with missing data, as commonly observed in ocean remote sensing data due to atmospheric conditions (e.g., cloud contamination). The performances of the proposed filter are tested and validated on the suppression of horizontal strip artifacts in cloudy L2 Sea Surface Temperature (SST) and ocean color snapshots.

scholarly journals The Use of Generalized Additive Model (GAM) To Assess Fish Abundance and Spatial Occupancy in North-West Bay of Bengal

2019 ◽  
pp. 17-28
Author(s):  
Bandanadam Swathi ◽  
Swarnalatha. V ◽  
Venkatesh Jogu
Keyword(s):  
Remote Sensing ◽  
Surface Temperature ◽  
Sea Surface Temperature ◽  
Generalized Additive Model ◽  
Chlorophyll Concentration ◽  
Additive Model ◽  
Remote Sensing Data ◽  
Sea Surface ◽  
Short Term ◽  
Sensing Data

The remote sensing data, such as sea surface temperature & chlorophyll concentration obtained from various satellites are utilized by Indian National Centre for Ocean Information Services (INCOIS) to provide Potential Fishing Zone (PFZ) advisories to the Indian fishing community which plays a vital role in national GDP. The data on Sea Surface Temperature (SST) is retrieved regularly from thermal-infrared channels of NOAA-AVHRR and chlorophyll concentration (CC) from optical bands of Oceansat-II and MODIS Aqua satellites for the identification of Potential Fishing Zones (PFZ) in Indian water. PFZ information has certain limitations, such as it can't predict the type of fish available in the notified fishing zone. In this dissertation, I have worked towards the development of short-term Hilsa shad predictive capabilities in a sustainable way. An effort has been taken to categorize all essential biological, environmental and climatic signals that have a direct or indirect impact on the Hilsa shad distribution. Remote sensing, ocean biogeochemical modelling, and statistical modelling approach have gained an increasing importance to study the marine ecosystems as-well-as for understanding the dynamics of the oceanic environment. Shad habitat has been studied from the geo-tagged fish catch data and oceanic/ecological indicators as predictor variables. For short-term prediction, the variables have been derived from a biophysical model, configured at INCOIS, using Regional Ocean Model System (ROMS) and remote sensing data. Using generalized additive model (GAM) Catch per Unit Effort (kg h?1) has been calculated as a response variable. Probability maps of predicted habitat with no fishing zone information have been generated using geographic information system.

Sign in / Sign up

Export Citation Format

Share Document