An NDVI-data harmonic analysis to study deforestation in Peru’s Tahuamanu province during 2001–2011

M. A. Muñoz Peña, F. A.R. Navarro

Research output: Contribution to journalArticlepeer-review

8 Scopus citations

Abstract

© 2016 Taylor & Francis. We investigate the spatio-temporal variability of normalized difference vegetation index (NDVI) associated with deforestation in four zones in southeastern Peru (Madre de Dios Region): (1) an area comprising part of the Tahuamanu province (mainly, but it also comprises a small part of Tambopata province), (2) a part of the Tahuamanu watershed, (3) surrounding areas of Iñapari and Iberia towns, and (4) another area in the Iñapari district which we labelled as Iñapari-B. We have utilized three statistical tools: the empirical orthogonal function (EOF), Fourier analysis, and wavelet analysis. EOF was used to obtain the maximum variability of the NDVI-anomaly spatial patterns and also their associated temporal variability as time series. Fourier analysis was used to determine the periodicities on the EOF pattern and associated them to deforestation in the region. Wavelet analysis was used to detect the maximum variability of the phenomena and to portray their statistical significance along the period of study and so identify when the maxima happen. From the results obtained, in Zone I, we have found a 2-month deforestation periodicity with major peaks in 2004 and 2009; in Zone II, the periodicity is about 1.8 months, with a maximum peak in 2009; in Zone III, the periodicity is 2.4 months with the highest peak in 2009; in Zone IV, the high-frequency periodicity is about 2.8 months with the main peaks in 2000, 2003, and 2004.
Original languageAmerican English
Pages (from-to)856-875
Number of pages20
JournalInternational Journal of Remote Sensing
DOIs
StatePublished - 16 Feb 2016

Fingerprint

Dive into the research topics of 'An NDVI-data harmonic analysis to study deforestation in Peru’s Tahuamanu province during 2001–2011'. Together they form a unique fingerprint.

Cite this