Assessment of climate variability by using Fourier components

• Published in: 1999 IEEE International Geoscience and Remote Sensing Symposium Vol. 1; pp. 681 - 683 vol.1
• Main Authors: Roerink, G.J., Menenti, M., Su, Z.
• Format: Conference Proceeding
• Language: English
• Published: IEEE 1999
• Subjects: Africa; Atmospheric modeling; Ecosystems; Europe; Image analysis; Image resolution; Predictive models; Satellites; Vegetation mapping; Weather forecasting
• ISBN: 0780352076; 9780780352070
• DOI: 10.1109/IGARSS.1999.773604

Vegetation is a very sensitive part of the ecosystem for climate change. Both the growing season and the total amount of vegetation are strongly affected by climatic changes. There is a strong need for more information about methods to identify and locate the climatic-sensitive regions, like desertification-threatened areas. To understand, identify and quantify the climatic effects on the regional ecosystems the relation between climate parameters, such as precipitation and net radiation, and the Normalized Difference Vegetation Index (NDVI) is studied and determined. The used climate parameters are the yearly precipitation and net radiation, which are output parameters of the weather prediction model RACMO (Regional Atmospheric Climate MOdel). The ratio of precipitation, P (converted from mm/d into W/m/sup 2/), over net radiation, R, (W/m/sup 2/), is called the climate indicator. In 1995 it ranged from zero in Northern Africa and some parts of Spain to values between 0.5 and 1 in Northern Europe to values larger than 1.5 in mountainous areas and parts of England. The NOAA/AVHRR satellite provides the 10-days-NDVI composites of 1995. The RACMO grid and the NDVI images cover most of Europe and the Mediterranean area with a resolution of respectively 0.5/spl deg/ and 0.01/spl deg/. In order to extract the characteristics of the vegetation cycle, a Fourier analysis is performed on the NDVI images. The Fourier components reflect the start, length and magnitude of the vegetation (=NDVI) growing cycle during the year. For this reason the Harmonic ANalysis of Time Series (HANTS) algorithm is developed.