Pixel-Based Model For High Latitude Dust Detection

Priftis, G ; Freitag, B ; Ramasubramanian, M ; Gurung, I ; Maskey, M ; Ramachandran, R

关键词: DUST;    DETECTION;    EARTH ALBEDO;    INFRARED RADIATION;    LIGHT (VISIBLE RADIATION);    MACHINE LEARNING;    MODIS (RADIOMETRY);    PIXELS;    POLAR REGIONS;    SPECTRAL SENSITIVITY;    WAVELENGTHS;   

RP-ID  :  MSFC-E-DAA-TN72807

学科分类：地球科学(综合)

美国|英语

来源: NASA Technical Reports Server

【 摘 要 】

Dust has implications on the energy budget, ocean biodiversity, and economy at regional and global scales. Dust detection relies on spectral sensitivity at visible (RGB) and infrared wavelengths. Radiative properties of high latitude dust and the background surface albedo in these regions (>40°N, >40°S) complicate current dust detection methods. Leveraging supervised machine learning (ML) methods, we propose a new method accounting for regional differences of dust occurrence.