Evaluation of vegetation cover of South Niger between 1984-2020
Satellite imagery analysis to identify the changes in vegetation over time in different seasons using ArcGIS
Reforestation is one of the widely recognized actions to mitigate climate change. The reforestation project Great Green Wall started in Africa in 2007. It had as its original goal to save land from the desert slowing down the encroachment of the desert into the Sahel region by reforesting an area originally 7,000 km long (east-west) and 15 km wide (north-south). Currently the project involves 20 countries stretching 8,000km across Africa.
This project aims to evaluate the regreening effort of part of the Zinder area over a period spanning 36 years using satellite images and vegetation index of both the dry and rainy seasons. To the best of my knowledge, there is not a study that used vegetation index to evaluate the land cover change of the area.
Data
The study area is located in the southern part of the Zinder region, Niger, (Lat 13° 00’53’’ N, Lon 009° 16’ 04” E) in the Easter Sahel. It covers an area of approximately 18,000 km2.
The satellite data used in the analysis included scenes collected by Landsat 4-5 TM, 7 ETM and 8 OLI/TIRS with WRS-2 coordinates (path 188 and row 051) at C2L2 processing level and cloud coverage <20%, from 1984 to 2020. When available, two scenes per year were retrieved, one at the end of the rainy season (September, October) and the other at the end of the dry season (February, March). No scenes matching these criteria were available from 1990 to 1999. Landsat 7 ETM scenes from 2003 to 2013 have missing data (~21%) because of the scan line correction (SLC) failure. The 2006, 2008 and 2012 scenes were analyzed without any modification or interpolation.
Workflow
To calculate the vegetation change over the years, MSAVI2 was chosen for the analysis since, compared to other indices, it produces a smaller "positive soil noise at low vegetation levels” (Jiang et al 2007) and it is recommended (Qi et al 1994) for areas with sparse vegetation and high degree of exposed and bright soil, whose light reflectance in the red and infra-red spectra could interfere with the vegetation’s. The formula used was:
which is derived from MSAVI =
but it doesn’t require specifying the soil brightness correction factor (L).
Once calculated, the index was reclassified in 4 classes: 1=<0; 2=0-0.33; 3=0.33-0.66; 4= > 0.66. Classes 3 and 4 were interpreted as new and established vegetation, however in the study area no cell belonging to class 4 were found
The scene was clipped to include only the Niger territory
Model used to calculate the MSVI2 and total surface in km2 for all the scenes. Link to the toolbox
Results
Overall, the MSAVI2 showed an increasing trend over the course of the years during the rainy season. Starting with less than 1% coverage in 1986 and approximately 5% vegetation coverage (class 3) in 2000, eventually a 53% coverage was observed at the end of the 2020 rainy season (Fig.1). At the same time, at a national level an increase in harvested area of all the major staple crops, such as cow peas, millet, sorghum, during the period 1984-2019 (FAOSTAT) has been observed (Fig.2).
As shown in Fig.3, the months of September and October correspond to the end of growing cycle and/or harvest time for the above-mentioned crops.
Fig.1 Change in vegetation cover at the end of the rainy season between 1984-2020
Between 2006 and 2012, 0% vegetation cover was observed. It is unclear what might have caused this. While a 21% missing data could explain at least part of the vegetation absence, it is also possible that the drought registered in 2005 (Goffner et al. 2019) would have had a significant impact. Indeed, the strong rainfall variability typical of the area may help explain the fluctuating crop yield (Dr. Alhassane, Global Yield Gap).
Fig.2 Area of major staple crops harvested between 1984-2019 in Niger
Fig.3 Niger's crop calendar
A significant difference in vegetation coverage was observed between the dry (February-March) and rainy (September-October) seasons. According to a vegetation survey (Sendzimir et al, 2011), Faidherbia albida is widespread in the region, and it loses its leaves during the rainy season. However, the analysis didn’t reveal any significant change in tree cover over the 36 years period during the dry season, when the tree species is supposed to have a full canopy. In that season, indeed, the vegetation (class 3 and 4) covered only 0.08% of the area even during 2020, the year in which the largest increase in green coverage was observed.
According to official statistics of the United Nation Convention to Combat Desertification (UNCCD), 146 million plants and seedlings were produced and over 720,000 ha of land was reforested and/or restored in Niger, including the southern part of the Zinder region, covered in this study. It is possible that the area selected in this study didn't include the area in which the regreening was more significant. Maybe there were other limitations due to image resolution, vegetation indicator or scene selected. More scenes should be included to cover a larger extent of the region to better understand the change that has occurred during the years.
References
United Nations, Convention to Combat Desertification https://www.unccd.int/actions/great-green-wall-initiative [retrieved on 9/13/2021]
Goffner et al., 2019. The Great Green Wall for the Sahara and the Sahel Initiative as an opportunity to enhance resilience in Sahelian landscapes and livelihoods. Reg Environ Change 19, 1417–1428
Sendzimir et al., 2011. Rebuilding resilience in the Sahel: regreening in the Maradi and Zinder regions of Niger. Ecology and Society 16(3):1
Modified Soil-Adjusted Vegetation Index 2 https://wiki.landscapetoolbox.org/doku.php/remote_sensing_methods:modified_soil-adjusted_vegetation_index
Jiang et al., 2007. Interpretation of the modified soil-adjusted vegetation index isolines in red-NIR reflectance space. Journal of Applied Remote Sensing
Qi et al., 1994. External factor consideration in vegetation index development. Proc. of Physical Measurements and Signatures in Remote Sensing, ISPRS, 723-730.
Niger, crop calendar. Reference Date: 05-January-2021 GIEWS - Global Information and Early Warning System. http://www.fao.org/giews/countrybrief/country.jsp?code=NER&lang=en FAO. [retrieved on 9/13/2021]
Global Yield Gap Atlas https://www.yieldgap.org/niger