Comparing Inter-Sensor NDVI for the Analysis of Horticulture Crops in South-Eastern Australia
Mohammad Abuzar,
Kathryn Sheffield,
Des Whitfield,
Mark O’Connell,
Andy McAllister
Issue:
Volume 2, Issue 1, February 2014
Pages:
1-9
Received:
24 February 2014
Published:
20 March 2014
Abstract: Synergistic application of NDVI from diverse sensors has been an interest of researchers in the field of natural resource management for over two decades. Attempts have been made to deal with the sensor-specific differences in NDVI which stem from a number of factors. In this study, an NDVI comparison has been made between Landsat-7 ETM+ and five other sensors of relatively fine resolution (ASTER, SPOT-5 XS, RapidEye, QuickBird-2 and WorldView-2) over an area of horticultural crops in south-eastern Australia during 2011-12. Translation equations have been developed using linear regression for specific sensors and specific horticultural crops (almond, table grape, wine grape, olive and vegetable). Cross-senor comparisons of NDVI showed strong positive relationships (p <0.001, R2>0.9) but in three cases (ASTER, SPOT-5 and RapidEye) the differences in NDVI values were significant (p<0.001) as well. Though in the other two cases (QuickBird-2 and WorldView-2) the differences were not significant, they were not negligible. Therefore the role of translation equations is considered important for cross-sensor NDVI compatibility. The results of this study will be used: (i) to convert NDVI from the selected sensors to a Landsat- equivalent NDVI for the analysis of irrigated horticultural crops, (ii) to optimise the temporal frequency of NDVI observations for long-term vegetation analyses, and (iii) to transfer Landsat ETM+-based measurements, particularly evapotranspiration (ET) estimates, to alternative sensors that lack thermal band capability which is critical for ET measurements. ET measurements will be used to estimate crop water requirement to help irrigation water management of horticultural crops.
Abstract: Synergistic application of NDVI from diverse sensors has been an interest of researchers in the field of natural resource management for over two decades. Attempts have been made to deal with the sensor-specific differences in NDVI which stem from a number of factors. In this study, an NDVI comparison has been made between Landsat-7 ETM+ and five o...
Show More
