Evaluation of Aridity Indices Using SPOT Normalized Difference Vegetation Index Values Calculated Over Different Time Frames on Iberian Rain-Fed Arable Land

Abstract

The objective of this study was to find the best-performing aridity index and time-frame in the Iberian Peninsula characterizing the effect of dryness on agricultural production. To achieve this goal time-series of 5 aridity indices for 1998 October–2009 December time-period were calculated on a 25 × 25 km grid, and the closest relationship with plant biomass was determined. Plant biomass was represented by the SPOT-VEGETATION Normalized Difference Vegetation Index (NDVI) satellite data masked out for rain-fed arable land for the period between 1998 and 2009, and also by official yield statistics of Spain and Portugal between 1999 and 2009. Aridity indices calculated for time frames matching the entire vegetative period resulted in the highest correlation coefficients with NDVI and with the crop yield. There was a difference between the two time frames covering twelve months. In contrast with the calendar year, using the hydrological year (1 October–30 September) ensured a very strong correlation between NDVI data and most aridity indices, with UNEP and Water Deficit aridity indices outperforming the others. Among the shorter time frames of April-October, January–October, and October–June, the latter provided very strong correlation between vegetation, UNEP and Water Deficit indices surpassing Budyko, De Martonne, and Thornthwaite aridity indices.

Keywords:

• aridity index
• Iberian Peninsula
• Normalized Difference Vegetation Index
• time frame

Acknowledgements

This work was partially supported by the Hungarian National Science Foundation (OTKA), Grant No. NN79835.

Notes

Note: “**” and “*” indicate correlation coefficients significant at 0.01 or 0.05 level, respectively.

Note: Yield is expressed in 100 kg/ha. NA = not applicable, the yield reported for the Calendar year was used.