Alpine Grassland Phenology as Seen in AVHRR, VEGETATION, and MODIS NDVI Time Series - a Comparison with In Situ Measurements - PubMed Skip to main page content
U.S. flag

An official website of the United States government

Dot gov

The .gov means it’s official.
Federal government websites often end in .gov or .mil. Before sharing sensitive information, make sure you’re on a federal government site.

Https

The site is secure.
The https:// ensures that you are connecting to the official website and that any information you provide is encrypted and transmitted securely.

Access keys NCBI Homepage MyNCBI Homepage Main Content Main Navigation
. 2008 Apr 23;8(4):2833-2853.
doi: 10.3390/s8042833.

Alpine Grassland Phenology as Seen in AVHRR, VEGETATION, and MODIS NDVI Time Series - a Comparison with In Situ Measurements

Fabio Fontana  1 , Christian Rixen  2 , Tobias Jonas  3 , Gabriel Aberegg  4 , Stefan Wunderle  5
Affiliations

Alpine Grassland Phenology as Seen in AVHRR, VEGETATION, and MODIS NDVI Time Series - a Comparison with In Situ Measurements

Fabio Fontana et al. Sensors (Basel). .

Abstract

This study evaluates the ability to track grassland growth phenology in the Swiss Alps with NOAA-16 Advanced Very High Resolution Radiometer (AVHRR) Normalized Difference Vegetation Index (NDVI) time series. Three growth parameters from 15 alpine and subalpine grassland sites were investigated between 2001 and 2005: Melt-Out (MO), Start Of Growth (SOG), and End Of Growth (EOG).We tried to estimate these phenological dates from yearly NDVI time series by identifying dates, where certain fractions (thresholds) of the maximum annual NDVI amplitude were crossed for the first time. For this purpose, the NDVI time series were smoothed using two commonly used approaches (Fourier adjustment or alternatively Savitzky-Golay filtering). Moreover, AVHRR NDVI time series were compared against data from the newer generation sensors SPOT VEGETATION and TERRA MODIS. All remote sensing NDVI time series were highly correlated with single point ground measurements and therefore accurately represented growth dynamics of alpine grassland. The newer generation sensors VGT and MODIS performed better than AVHRR, however, differences were minor. Thresholds for the determination of MO, SOG, and EOG were similar across sensors and smoothing methods, which demonstrated the robustness of the results. For our purpose, the Fourier adjustment algorithm created better NDVI time series than the Savitzky-Golay filter, since latter appeared to be more sensitive to noisy NDVI time series. Findings show that the application of various thresholds to NDVI time series allows the observation of the temporal progression of vegetation growth at the selected sites with high consistency. Hence, we believe that our study helps to better understand largescale vegetation growth dynamics above the tree line in the European Alps.

Keywords: AVHRR; NDVI.; grassland; phenology; remote sensing.

PubMed Disclaimer

Figures

Figure 1.
Figure 1.
Spatial distribution of the sites (black dots) in Switzerland. All sites represent subalpine and alpine grassland (modified from [31]). The numbers associated with the black dots indicate site elevations above sea level [m].
Figure 2.
Figure 2.
Sample data (black dots) from the ultrasonic sensor at the Tujetsch site (2270 m a.s.l.) in 2001. Melt-out (IMISMO), start of growth (IMISSOG), and end of growth (IMISEOG) are determined from a 3-leg linear fit of the growth signal (adapted and modified from [31]).
Figure 3.
Figure 3.
Composite NDVI time series (AVHRRMVC; dotted) at the Dötra site in southern Switzerland (2060 m a.s.l) in 2002 and the corresponding Fourier adjusted (thin solid) as well as Savitzky-Golay filtered NDVI products (dashed). NDVI increase in spring is very pronounced after snow melt. The Savitzky-Golay product follows AVHRRMVC more closely compared to the Fourier product. Note the temporal offset between AVHRRMVC (based on precise acquisition dates) and the Savitzky-Golay NDVI product (fix time steps of i=10 days assumed). Thick solid lines mark the thresholds (th) where 50% (th=0.5), 75% (th=0.75), and 98% (th=0.98), respectively, of total annual NDVI amplitude are crossed for the first time.
Figure 4.
Figure 4.
Relationship of remote sensing NDVI time series and IMIS ground data depending on the chosen threshold (th), exemplary shown for the detection of melt-out (MO) by means of the Fourier adjusted MODIS (500 m) NDVI data set. Where th is chosen too low (th=0.4; left) the majority of the points lies below the 1:1 line (mean offset of days (OD¯) negative). OD for each point is given by the vertical distance between each point and the 1:1 line. With th=0.5 (middle) points closely follow the 1:1 line (OD¯≈0), whereas the point cloud shifts over the 1:1 line (OD¯positive) with increasing thresholds (th=0.6; right). r=linear correlation coefficient of satellite and ground data set, σ=standard deviation of OD.
See this image and copyright information in PMC

Similar articles

See all similar articles

Cited by

See all "Cited by" articles

References

    1. Studer S., Stöckli R., Appenzeller C., Vidale P. A comparative study of satellite and ground-based phenology. International Journal of Biometeorology. 2007;51:405–414. - PubMed
    1. Menzel A. Trends in phenological phases in Europe between 1951 and 1996. International Journal of Biometeorology. 2000;44:76–81. - PubMed
    1. Roetzer T., Wittenzeller M., Haeckel H., Nekovar J. Phenology in central Europe - differences and trends of spring phenophases in urban and rural areas. International Journal of Biometeorology. 2000;44:60–66. - PubMed
    1. Defila C., Clot B. Phytophenological trends in Switzerland. International Journal of Biometeorology. 2001;45:203–207. - PubMed
    1. Myneni R.B., Keeling C.D., Tucker C.J., Asrar G., Nemani R.R. Increased plant growth in the northern high latitudes from 1981 to 1991. Nature. 1997;386:698–702.

LinkOut - more resources