Performance tests to modeling future climate-vegetation interactions in virtual world: an option for application of remote sensed and statistical systems

Azeddine Hachmi, Asmae Zbiri, Dominique Haesen, Fatima Ezzahrae El Alaoui-Faris, David A. Vaccari

Research output: Contribution to journalArticlepeer-review

2 Scopus citations

Abstract

Working in the virtual world is different to real experiment in field. Nowadays, with remote sensing and new analysis programs we can assure a quick response and with less costs. The problem is efficiency of these methods and formulation of an exact response with low errors to manage an environmental risk. The objective of this article is to ask question about performance of some tools in this decision making in Morocco. The study uses (Test 1: TaylorFit Multivariate Polynomial Regressions (MPR); Test 2: SAS Neural Network (NN) to modeling relationship between European Center for Medium-Range Weather Forecasts dataset and NDVI eMODIS-TERRA at arid Eastern Morocco. The results revealed that the both test could accurately predict future scenario of water stress and livstock production decrease. The experience shows that virtual work with Artificial Intelligence is the future of ecological modeling and rapid decision-making in case of natural disasters.

Original languageEnglish
Pages (from-to)178-189
Number of pages12
JournalWSEAS Transactions on Information Science and Applications
Volume18
DOIs
StatePublished - 2021

Keywords

  • Decision making
  • ECMWF Rainfall
  • Eastern Morocco
  • Multivariate Polynomial Regressions (MPR); Neural Network (NN)
  • NDVI
  • Performance Tests
  • Virtual world

Fingerprint

Dive into the research topics of 'Performance tests to modeling future climate-vegetation interactions in virtual world: an option for application of remote sensed and statistical systems'. Together they form a unique fingerprint.

Cite this