loading page

Urban Air Temperature Model Using GOES-16 LST and a Diurnal Regressive NeuralNetwork Algorithm
  • +1
  • Joshua Hrisko,
  • Prathap Ramamurthy,
  • Yunyue Yu,
  • Peng Yu
Joshua Hrisko
CUNY City College of New York

Corresponding Author:jhrisko000@citymail.cuny.edu

Author Profile
Prathap Ramamurthy
Princeton University
Author Profile
Yunyue Yu
NOAA
Author Profile
Peng Yu
NOAA National Environmental Satellite, Data, and Information Service
Author Profile

Abstract

An urban air temperature model is presented using GOES-16 land surface temperature. The Automated Surface Observing System (ASOS) serves as ground truth air temperature for calibration and testing of the model. The National Land Cover Database (NLCD) is used to calculate a weighted distribution of 20 land classifications for each satellite pixel surrounding a nearby ASOS station. A time-match algorithm aligns the ground and satellite measurements within 5-minutes of one another, and the resulting matched LST and air temperature are compared over nine months to investigate their cross-correlation. A model is constructed by fitting their difference using a gaussian profile. Landcover, latitude, longitude, local time, and elevation are inputted into an artificial regressive neural network to fit each unique GOES-16 pixel. Over 100 urban stations and satellite pixels throughout the continental U.S. are used to construct the diurnal gaussian model and approximate air temperature. Early statistics indicate favorable results, competing with other studies with more complicated and intensive calculations. The presentation of this model is intended to simplify the calculation of air temperature from satellite LST and create a successful model that performs well in urban environments. The improvement of urban air temperature calculations will also result in improved satellite land surface products such as relative humidity and heat index.