Navigation – Plan du site

AccueilRubriquesCartographie, Imagerie, SIG2016Detecting Changes in Vegetation T...

2016
779

Detecting Changes in Vegetation Trends in the Middle East and North Africa (MENA) Region Using SPOT Vegetation

Détection des changements de la végétation dans la région du Moyen-Orient et de l’Afrique du Nord (MENA) en utilisant le SPOT Végétation
Ghaleb Faour, Mario Mhawej et Abbas Fayad

Résumés

La région du Moyen-Orient et de l’Afrique du Nord (MENA) peut être considérée comme la région la plus pauvre en eau du monde, affectant ainsi l’état de la végétation dans cette zone. Des données du Normalized Difference Vegetation Index (NDVI), qui s’étalent temporellement, sont utilisées dans le monde entier pour la cartographie et la surveillance de grande surface de végétation. Ils sont souvent à base de la télédétection. Cette étude évalue le changement de la végétation dans les pays arabes en utilisant le SPOT Végétation qui calcule le NDVI entre 1999 et 2012. Cinq catégories ont été identifiées: Hot spot ou changement négatif important, changement négatif, pas de changement, changement positif et Bright spot ou changement positif important. Les résultats indiquent une diminution importante du couvert végétal dans environ 553 000 km2 de la surface totale de la région arabe. D'autre part, seulement moins de 1% de la région a connu des changements positifs. Les zones ayant des changements de végétation positives sont principalement situées en Algérie et en Egypte, ainsi que dans la partie sud de la Somalie et l'Irak. Les changements négatifs ont été visibles presque partout dans MENA, indépendamment des pays, du climat et de la topographie. On constate que les changements de la végétation dans cette zone sont principalement liés aux activités humaines; le changement climatique ne joue qu'un rôle secondaire dans la région MENA.

Haut de page

Texte intégral

Introduction

1Vegetation dynamics reflect the net effects of many factors, including climate, abiotic environment, biotic interactions, and disturbance history. Understanding how these factors interact and affect species coexistence and productivity is crucial to build future strategies and mitigations. The Middle East and North Africa region (MENA) is under severe vegetation changes. According to the Intergovernmental Panel on Climate Change (IPCC, 2008; Terink, 2013) report, precipitation will decrease and temperature will increase consequently across the MENA over the next decades and eventually affecting evaporation and soil moisture content. The spatial changes in the amount, intensity, and frequency of these climate indicators shall affect the magnitude and frequency of vegetation dynamics.

2The Normalized Difference Vegetation Index (NDVI) is one of the mostly used indexes worldwide to identify and measure the vegetation dynamic. It quantifies the amount of green vegetation. Studying its trend could generate an indication on the vegetation changes in an area: positive trend reflects an increase in the vegetation cover, in term of density and/or area, whereas negative trend show a greenness decrease. Nevertheless, the negative NDVI trend isn’t only related to vegetation degradation, and it can be caused by different factors, including precipitation variability, changes in land management input, poor sensor calibration, orbital drifts, etc. The combination of vegetation dynamic models and remote sensing time-series images is being used for the assessment and prediction of temporal variations in canopy cover, primary production, crop yield, crop growth, and crop production (e.g. Gillies and Carlson, 1995; Haboudane et al., 2004; Donohue et al., 2014; Ardö, 2015; Du et al., 2016). Actually, GIS and allied technologies, namely Remote Sensing, contributions in the field of vegetation exploitation and forecasting are not only due to their capabilities to support spatial dimension but also to their ability to capture, store, organize, process, analyze, and create outputs. Global NDVI data are routinely derived from three type of satellites: the AVHRR (Advanced Very High Resolution Radiometer), SPOT-VGT (Satellite for observation of Earth), and MODIS (Moderate Resolution Imaging Spectroradiometer) earth observation records (e.g. Tucker et al., 2005; Brown et al., 2006; Tarnavsky et al., 2008; Fensholt, 2012; Los, 2013; Tian et al., 2015). While AVHRR datasets contain historical data record at coarser resolution of 8 km; SPOT-VGT and MODIS are generally used in short term study with a resolution of 1 km and 250 m., respectively. Actually, time-series Normalized Difference Vegetation Index (NDVI) datasets are being used for mapping and monitoring large areas (Neeti and Eastman 2001; Eastman et al., 2009; Li et al., 2013; Tian et al., 2013; Luo et al., 2013; White et al., 2014). The main advantages of NDVI are basically associated with its global coverage, high temporal resolution (daily to monthly), medium spatial resolution (1 to 10 km), its efficiency in processing time and storage capacity, and its availability free of charge.

3Time series analysis requires improved techniques and up-to-date remote sensing datasets to support the assessment of vegetation degradation at a regional scale. However, programs using NDVI time series as an input in describing, assessing, and monitoring vegetation dynamics are most of the time oriented towards quantifying spectral characteristics of the earth’s surface. The applications include phenological analysis of vegetation cover (Zhou et al., 2001; Stockli and Vidale, 2004; Julien and Sobrino, 2004; Walker and Wynne, 2014), land-cover mapping and classification (Townshend, 1994; Friedl et al., 2002; Tucker et al., 2005; Alves et al., 2015) crop yield forecasting (Seiler et al., 2000, 2007; Kouadio et al., 2014; Shao et al., 2015).

4The assessment of time series datasets is less appraised. Actually, time series analysis is used for capturing seasonal and inter-seasonal (Funk et al., 2006; Chen et al., 2015) and annual variability (Rasmussen et al., 1997; Pouliot et al., 2014). Concerning the MENA region, several literatures study vegetation changes in various parts of our study area. These detect environmental changes, mainly in the southern (Jabbar and Zhou, 2013) and central part of Iraq (Othman et al., 2014; Becker, 2014), assess NDVI Spatial Patterns in Saudi Arabia (Aldakheel, 2011), evaluate rangeland degradation in Syria (Geerken and Ilaiwi, 2004; Udelhoven and Hill, 2009) study the impact of the intense drought of 2007-2009 over the Fertile Crescent area (Trigo et al., 2010), forecasts yields in the North African countries (Maselli and Rembold, 2001) analyses degradation of steppe vegetation in Algeria (Khader et al., 2014), and monitor desertification in South Tunisia (Elmzoughi et al., 2008).

5In this study, NDVI derived from SPOT Vegetation over the time period between 1999 and 2012 are used. The need to establish a comprehensive monitoring system for the quantitative evaluation of vegetation degradation is the main motivation behind this research. SPOT NDVI (1999 – 2012) provides information on the recent changes in the vegetation cover. Short-term monitoring allows for capturing the most recent forcing stress or pressure factors, namely drought, urban development, overuse of the non-renewable water resources, and natural hazards and extremes. The main objective is to establish a method that emphasizes on the use of remotely sensed time series data. In addition, an extended summary of pilot areas as an input to describe, assess, and monitor vegetation dynamics and to evaluate vegetation degradation propagation at regional scales are also present.

6The specific objectives of this research focus on: (i) assessing vegetation dynamics and trends from remote sensing time series NDVI; (ii) classifying the study zone into different classes ranging from the most severe to the most developed vegetation covers; (iii) identifying critical hot spot areas that need a fast action over the MENA region, which is one of the most arid and water scarcest places in the world.

Study Area and Data

Study Area

7The MENA region in this paper refers to the territories of the Arab countries (Figure 1). These Countries are member in the Arab league, and more specifically, Arab Center for the Studies of Arid Zones and Dry Lands (ACSAD).This region covers around 14 million square kilometers, which is around 9 per cent of the Earth’s total land area, with a population of over 350 million people, of which 56% are urbanized.

8According to the Köppen Climate Classification System the MENA region is mostly dominated by hot and arid climate with exceptions that exist in the coastal areas and highlands. The North African and West Asia regions are dominated by the Desert climate where the Sahara Desert covers 90 per cent of the total North African region. The prevailing temperature usually exceeds the 50°C and the precipitation annual average is less that 25 mm. The Arid Steppe climate with cold winters prevails in Morocco, Algeria and Tunisia, except the region of the Atlas Mountains where the climate is cooler and precipitation is higher. Regions including the southern part of Sudan and Somalia are characterized by the Equatorial climate where average precipitation exceeds 1500 mm annually, and the average temperature is around 20°C. The Warm Temperate climate (i.e. Mediterranean Climate) dominates the coastal North Africa and the eastern Mediterranean countries. The Mediterranean climate is characterized by mild to cool, wet winters with hot, dry summers. Annual average temperature ranges between 7 and 25°C and the annual precipitation is around 520 mm. the inland areas of Syria, Jordan and Iraq are dominated by the Arid Steppe Climate. The arid to hyper-arid climate prevails in the Arabian Peninsula, where the average annual precipitation does not exceed the 150 mm and the average temperatures varies drastically between winter and summer seasons (ranges between 5 to 15°C during winter and between 40 and 50°C during summer).

Figure 1: Territories of the Arab countries

Figure 1: Territories of the Arab countries

9Of the total 14.1 million km2 land area, 89.3% is arid or hyper-arid land. Agricultural lands cover around 2 million km2 (~14.5 per cent of the total area), of which only 4.2% is cultivated. Forest areas are largely limited to Sudan (~62000 km2) and Somalia (~7500 km2) and to less extent in regions of Morocco and Algeria and the coast of the western Mediterranean countries. Nevertheless, the poor management, agricultural practices and increased deforestation significantly exacerbate land degradation.

10Based on the past 30 year trends, it is expected that the MENA region shall continue to suffer from severe degradation of lands resulting from anthropogenic drivers and namely from deforestation, agricultural and industrial pollution, and the increase in urban and infrastructural development at the expense of agricultural lands. Natural hazards are also present to further contribute to land degradation in the MENA region. Currently, drought is the most prevailing natural hazards in the MENA region. Extreme temperatures, wildfires, flooding, landslides, and sand and dust storms are also among the natural events that negatively impact land and contribute to land degradation. In addition, climate changes exhibit a further impact on vegetation throughout the region. Although the countries in the region share common concerns about a number of critical environmental and sustainability issues, they also face tremendous differences in their challenges.

SPOT VEGETATION NDVI Data

11This study was performed by using NDVI data derived from the SPOT VEGETATION sensor (VGT) on board the SPOT4 and SPOT5 satellites with a spatial resolution of 1 km from 1999 to 2012. These data are accessible free of charge at (http://www.vito-eodata.be/​). VGT-S10 data (10-day synthesis product) are generated by selecting the pixels that have the maximum NDVI values within a 10-days period. The NDVI downloaded dataset was interpreted to provide 14-years satellite record of monthly changes in terrestrial vegetation. Maximum Value Compositing (MVC) of the NDVl datasets has been employed using ERDAS Imagine 2013 to minimize cloud contamination.

Reference Datasets

12Reference data were based on Lebanese CNRS (National Council for Scientific Research) and ACSAD (Arab Center for the Studies of Arid Zones and Dry Lands) satellite archive database covering the main Arab countries. This database is based on the NASA earth observatory (http://earthobservatory.nasa.gov/​), Landsat TM and ETM+ (http://landsat.usgs.gov/​) and the MODIS Land use and Land cover data products (MOD12 datasets available at http://reverb.echo.nasa.gov/​) between 2000 and 2013. In addition, several vegetation changes areas were collected from regional studies and reports (e.g. Afifi et al., 2010; Daccache et al., 2014).

Methods

Trend vegetation changes

13Parametric and non-parametric approaches have been commonly used for NDVI trend estimation purposes. However, parametric trend results might not be reliable when assumptions about normality are not met. In this context, different tests of normality have been developed and can be found in literature. One of these tests is the Jarque-Bera (JB) test (Jamali et al., 2012). The JB measures the conformance of a variable in relation with a normal distribution.

14Jarque-Bera test was applied to the NDVI time series dataset (SPOT 1999-2012) with a significance level of 95% (α = 0.05). The output map shows that more than 80% of the land is not following a normal distribution. Consequently, it is necessary to combine both parametric and non-parametric approaches.

15As a result, NDVI time series data were estimated by combining two approaches: Ordinary Least-Squares (OLS) regression, applied in the region with a normal distribution, and the combined Mann-Kendall test with the Sen’s Slope estimator, applied in the region with non-normal distribution.

16NDVI changes map between 1999 and 2012 were produced. However, in aim to highlight only the vegetated areas in 1999, we used the Global Land Cover 2000 map already produced by FAO (Food and Agriculture Organization). It was a collaboration of 30 research teams from around the world. GLC2000, which is freely accessed online at http://www.glcn.org/​activities/​glc2000_en.jsp, compiled harmonized global land cover classification for the year 2000, using data acquired by the SPOT4 Vegetation instrument. The merging between NDVI changes map and GLC2000 enables to create a Trend Vegetation map containing only vegetated areas from 2000. Thus, we eliminate the instrument errors generated from non-vegetated areas such as bare land, sparse vegetation, etc. These areas fall under the “no change” class.

Classification

17The NDVI trend maps are categorized in five different classes, ranging from hot to bright spot as follows:

  • Hot spot: suffering from forest fires, high precipitation anomalies, intensive drought or recurring dry years, negative vegetation anomalies, specific local or regional factors (such as wars) with a vegetation decrease of more than -15%;

  • Negative change: subject to recurring moderate drought, moderate vegetation, precipitation, and temperature anomalies, urban expansion with a vegetation decrease between -15% and -5%;

  • No change: urban and desert areas with ±5% change in vegetation values

  • Positive change: containing positive precipitation and vegetation changes, and agricultural areas near flooded zones with a vegetation increase between 5% and 15%;

  • Bright spot: containing development of agricultural schemes, high positive precipitation trends and multi-years positive vegetation anomalies; vegetation increase more than 15%.

18These classifications have been applied as the standard threshold since 2007 in the biennial bulletin (CNRS, ACSAD and GIZ, 2015) for land degradation monitoring and assessment in the Arab world.

Results

Vegetation changes in the MENA area (1999-2012)

19Results obtained over the time period between 1999 and 2012 by the analysis of the SPOT Vegetation Products reveal that the vegetation degradation accounted for nearly 553 000 km2, thus 4.67% of the Arab world (Table 1 and Figure 2). These areas exist mainly in the western part of Morocco, the northern part of Syria and Somalia, and in southern part of Sudan. Meanwhile, little improvement in the vegetation cover occurred with an estimated development of 0.41% of the total Arab area, irrespective of the location, climate and topography (Table 1 and Figure 2). Positive changes are mostly located in the northern part of Algeria and Egypt, as well as in the southern part of Somalia and Iraq.

Table 1: Classified vegetation changes in different Arab countries in km2, 1999-2012

Arab Countries

Hot Spot (%)

Negative Change (%)

No change (%)

Positive Change (%)

Bright Spot (%)

Algeria

0.26

0.25

99.29

0.17

0.02

Bahrain

0.00

0.00

100.00

0.00

0.00

Djibouti

0.06

5.06

94.38

0.25

0.24

Egypt

0.03

1.33

98.52

0.08

0.04

Iraq

2.59

12.70

83.34

0.85

0.53

Jordan

0.04

0.52

99.40

0.03

0.01

Kuwait

0.00

0.69

99.28

0.03

0.00

Lebanon

0.57

1.12

97.98

0.32

0.02

Libya

0.02

0.83

99.12

0.03

0.00

Mauritania

0.03

1.13

98.79

0.05

0.00

Morocco

12.00

2.58

85.15

0.26

0.01

Oman

0.01

0.31

99.67

0.01

0.00

Qatar

0.00

0.41

99.59

0.00

0.00

Saudi Arabia

0.11

0.54

99.26

0.06

0.11

Somalia

0.18

10.63

83.94

2.36

0.18

Sudan

0.24

2.37

97.13

0.16

0.24

Syria

3.61

4.96

91.38

0.05

3.61

Tunisia

0.13

1.24

98.20

0.39

0.13

UAE

0.02

0.47

99.43

0.06

0.01

Western Sahara

0.00

0.01

99.99

0.00

0.00

Yemen

0.21

0.82

98.95

0.01

0.00

Total

0.93

3.74

90.56

0.24

0.17

Figure 2: Areas with vegetation changes based on the SPOT Vegetation NDVI dataset, 1999-2012

Figure 2: Areas with vegetation changes based on the SPOT Vegetation NDVI dataset, 1999-2012

20According to the Koppen classification, the MENA region is classified into four climatic zones (i.e. Arid, Equatorial, Snow, Warm temperature). While no change area has the largest percentage in these four classes, negative change and hot spot area came second with a percentage ranging between 12.43% in thewarm temperature zone to 0.04% in the snow areas. As a result, warm temperature area is the most affected region, causing a high rate of vegetation degradation. However, positive change to bright spot ranges between 0.41% in the arid region and 0.31% in the warm temperature zone, with no positive changes observed in the snow areas. The arid region displays the largest positive change in the vegetation cover (Table 2).

Table 2: Percentage of vegetation change in different climatic zones within the Arab region, 1999-2012

Climatic zones

Hot spot

Negative change

No change

Positive change

Bright spot

Warm temperature

8.04

4.40

87.17

0.30

0.09

Arid

0.39

1.98

97.23

0.24

0.18

Snow

0.04

0.00

99.96

0.00

0.00

Equatorial

0.49

0.06

99.14

0.04

0.27

Accuracy Assessment

21Pilot areas (350 sites) were used to validate the model output and the accuracy of the generated land degradation maps. Overall accuracy, producer’s accuracy, as well as user’s accuracy was used to quantitatively determine how effectively pixels from the land degradation map were grouped into the correct status of the pilot areas. These GCP (Ground Control Point) was created from the Lebanese CNRS-ACSAD archive database, as well as from regional studies and reports. These locations represent areas where events have occurred, such as drought, flood, fires, expansion in agriculture, stabilizing and improved agriculture as well as the relatively stable area (i.e. mainly desert zones). Pilot areas identified changes that occurred between 2000 and 2012 and are only used in the validation for the SPOT derived land degradation maps. For each pilot area, a rank has been given according to evidence from different sources to assume a direct or indirect causality among the changes in the vegetation productivity and environmental alterations.

22A confusion matrix is presented in Table 3, showing the comparison between map trend and ground truth data. The overall accuracy of the proposed methodology is estimated at 71.11%. With negative and positive changes based on literature not well represented in our model generated classes (i.e. 37.20% and 52.83% of producer’s accuracy respectively), more reference data that include these categories are recommended. In fact, most published scientific research papers focused on areas with extreme changes (i.e. hot spot and bright spot) in the vegetation cover, and not on moderately changing areas. However, the average producer’s accuracy as well as the average user’s accuracy are 68.83% and 69.43% respectively.

Table 3: Contingency table between the land classification system and GCP (Ground Control Point) from CNRS-ACSAD database and literature

Map data

Reference Data

Hot Spot

Negative Change

No Change

Positive Change

Bright Spot

SUM

User's Accuracy

Hot Spot

54

15

9

0

0

78

69.23%

Negative Change

4

16

6

2

0

28

57.14%

No Change

0

10

69

4

0

83

83.13%

Positive Change

0

1

8

28

10

47

59.57%

Bright Spot

0

1

5

19

89

114

78.07%

SUM

58

43

97

53

99

350

69.43%

Producer's Accuracy

93.10%

37.20%

71.13%

52.83%

89.89%

68.83%

Overall Accuracy

(54+16+69+28+89)/350 = 71.11%

Discussion

23According to TRMM Online Visualization and Analysis System (TOVAS), accessible online at http://giovanni.gsfc.nasa.gov/​, the rainfall regime in the North Morocco from 1999 to 2012 has increased by 37% as it is shown in Figure 3. While an increase in the vegetation cover is expected, North Morocco area has an area of nearly 65106 km2 of negative changes in the vegetation cover – 14.58% of Morocco total area. As a result, the amount of precipitation in this region does not affect the observable process of degradation. In fact, the western part of Morocco, namely Agadir and Essaouira regions, is facing many pressures that accelerate land degradation in the region. Maintaining too many animals by extensive use of concentrates has recently led to a large scale degradation of the vegetation cover (Berkat and Tazi). In addition, mineral extraction and quarrying exceeded 1 million tons a year since 2003, leading to severe land degradation. Agadir region has an average annual rainfall of 250 mm., which is almost the same as Essaouira area (i.e. 278 mm.) (Berkat and Tazi, 2004). However, Essaouira, unlike Agadir, is still classified as agro-pastoral region. Arable land expansion and overgrazing led to deforestation, which causes water and wind erosion. Coupled with drought, topography, soil and vegetation overuse, as well as current intensive farming systems; they contribute in the desertification process in prone semi-arid zones such as the Essaouira area. Such disruptions affect the biological resources; and therefore, the lives of the people who end up leaving their fields to move to neighboring cities or emigrate, which end up exacerbating desertification process.

24According to TRMM Online Visualization and Analysis System (TOVAS), the rainfall regime in the North Algeria and Tunisia from 1999 to 2012 has increased by 28% as it is shown in Figure 4. An increase in the vegetation cover is observable in this region. In addition to the precipitation effect, in the inner lands, irrigation helped in developing new green spaces, especially in the Algerian high plains (i.e. Sétif) and the coastal area of Tunisia – i.e. expansion of the olive groves of Sfax, citrus in Nabeul, etc. Furthermore, Algerian government attempts major renovations in planning, policies and attitudes to overcome the water scarcity. The first dam was built in 2009, involving the transfer of 155 million m3 of water per annum, serving a population of 360000 inhabitants. Thirty new dams with a total capacity of nearly 1.5 billion m3 will be realized in the future five-year plan from 2015 to 2019. This attitude, added to a significant increase in the precipitation regime, has developed the vegetation regions; thus NDVI shows a positive to bright spot in the northern part of Algeria. However, Algeria and Tunisia are experiencing large scale urbanization over their coastal plains leading to a concentration of population and activities in this zone at the expense of agricultural areas.

25According to TRMM Online Visualization and Analysis System (TOVAS), the rainfall regime in the South Iraq from 1999 to 2012 has decreased by 14% as it is shown in Figure 5. This region represents positive and negative vegetation change. In the Euphrates valley of Iraq, the vegetation index is increasing due to the expansion of irrigation practices as well. Meanwhile, vegetation degradation in the southern part of Iraq was accelerated by the draining of the Central and Hammar marshes during the previous political rules in order to evict the rebels. The Glory River was also constructed to divert water from the marshes. As a result, from 1991 to 2003, this area showed deterioration in its vegetation cover resulting from the loss of around 90% of its water resources. Nonetheless, due to water restoration, this area is starting to show bright development in its vegetation cover. Still, the rainfall regime does affect many part of this region resulting in a large vegetation degradation. Syria, on the other hand, has been overwhelmed by a civil war that severely affected not only the country but also the ecosystem. As a result, Syria has been suffering from a decrease in the vegetation cover since 2011. This impact will be easily identified in the coming years, as more degradation is expected as a consequence of the on-going civil war.

26According to TRMM Online Visualization and Analysis System (TOVAS), the rainfall regime in the North West Saudi Arabia from 1999 to 2012 has decreased by 44% as it is shown in Figure 6. While a decrease in the vegetation cover is expected, the Arabian Peninsula witnessed several developments in irrigation practices in the desert. It is due to the exploitation of the groundwater resources, mainly aided by the government in these countries. Over the past decade, the agriculture in the Saudi Arabia has drastically improved. Saudi Arabia is converting large areas of desert into agricultural fields by implementing major irrigation projects and adopting large scale mechanization. Although, this positive development is nearly shown in the northern part of Saudi Arabia. In fact, due to low economic water productivity, Saudi Arabia adjusted policy towards investing in agriculture in areas with fewer hostile climates and renewable water resources.

27Armed with its abundance wealth and political connections, Saudi Arabia has invested around 11 billion USD in agricultural and livestock projects abroad. Somalia was one of those countries that benefit from this policy. They believe that Africa is the region that represents the biggest opportunity to increase food production with vast tracts of land and a big difference between existing potential and current productivity. It also encourages the Saudi private sector to invest abroad using their financial surpluses, their agriculture experience and the modern production and irrigation technology they have. As a result, Somalia is presenting a positive to bright spot in 16196 km2, mainly due to the investment in their agricultural fields. According to TRMM Online Visualization and Analysis System (TOVAS), the rainfall regime in the South Somalia from 1999 to 2012 has increased by 22% as it is shown in Figure 7. Thus, the precipitation in this region is a helping factor for the vegetation cover’s increase. A further discussion on the amount of improvements from the precipitation regime is required. In contrast, the rainfall regime in the North Somalia from 1999 to 2012 has increased by 21% as it is shown in Figure 8. While an increase in the vegetation cover is expected, North Somalia area represents generally a degraded vegetation cover. Floods are the major cause for this degradation (Artan et al., 2007).

28The rainfall regime in the Middle Sudan from 1999 to 2012 has decreased by 10% as it is shown in Figure 9. As result, the regression in the vegetation degradation is accounted for the decrease of the rainfall regime, added to an anthropic factor, such as urbanization and overexploitation. Further studies in this region should be held in a way to quantify each interfering factor. Meanwhile, the rainfall regime in the South Middle Sudan, named as The Sudd, from 1999 to 2012 has decreased by 18% as it is shown in Figure 10. This region is a vast swamp in South Sudan, formed by the White Nile's Bahr al-Jabal section. Added to the regression of the precipitation regime, the Sudd, especially in Malakal area, has been destroyed by the war. The city that before the civil war was inhabited by 250,000 people remains a military target. The vegetation is now destroyed at the point that you can no longer make out the landmarks that existed before the destruction of the area.

29The rainfall regime in Lebanon and Jordan from 1999 to 2012 has decreased by 6% as it is shown in Figure 11. Across Lebanon and Jordan, there is a negative change in the vegetation cover due to the dryer climat, the pressure on water resources in addition to the intense urbanization of these already densely populated regions within the Arab world.

30In conclusion, vegetation changes over the MENA region is mainly caused by human activities and theirs policies (e.g. civil wars, extensive use of concentrates, arable land expansion, overgrazing, urban expansion, irrigation practices, investments in agriculture, etc.). Climate change plays a secondary role in these changes. Future plans and strategies shall consider the impact of human interventions as a primary contributor in this phenomenon to succeed any mitigation initiatives and respond to a sustainable development of the considered region.

31Monthly VGT-S10 data product has decreased the detection of vegetation in arid and Sahara area due to the presence of a short term vegetation cover, generally undetected by satellite sensors on a monthly based scale. Therefore, higher temporal and radiometric resolutions products are recommended. In addition, NDVI is influenced by the soil type and land cover classifications, specifically in the low vegetation areas. Using a specialized index for soil, e.g. SAVI (Soil-adjusted Vegetation Index) and PVI (Perpendicular Vegetation Index) shall certainly increase the accuracy in these regions, by eliminating artefacts and normalizing soil effects (Huete and Tucker, 1991). While long term studies are recommended, our year by year data processing and interpretation for our region shall achieve that goal in the coming years.

32Studies (Bai el al., 2008; Rasmussen et al., 2006; Fensholt et al., 2013) focused on land degradation in a large scale, used only a linear regression model. However, the combination of both linear and non-linear methods, for a recent time period (i.e. 1999-2012) represents better our diverse study area. Even though an addition of the climatic data factors shall adjust results positively, the lack of sufficient climatic station can interrupt the applicability of this method in the region. Because of the extremely large study area, more GCP (Ground Control Points) are advised to eliminate any bias of the areas’ estimates and their classifications. The lack of available data in this region is the main issue though. Nonetheless, TRMM (Tropical Rainfall Measuring Mission) and GPM (Global Precipitation Measurement) datasets can provide next-generation global observations of precipitation from space. Besides, land degradation can be caused by different factors, including climate change, government policies, anthropogenic dynamics, etc. (Evans and Evans, 2004). While government policies and anthropogenic factors exacerbate land degradation, climate change is showing a minor contribution to this phenomenon – e.g. drought was from the main cause of the civil war in Syria (Balanche, 2013). However, future papers that illustrate the relation between these factors are vital in assessing land degradation, while providing essential tools for the decision makers.

Figure 3: Rain rate trend in North Morocco, 1999-2012

Figure 3: Rain rate trend in North Morocco, 1999-2012

Figure 4: Rain rate trend in North Algeria and Tunisia, 1999-2012

Figure 4: Rain rate trend in North Algeria and Tunisia, 1999-2012

Figure 5: Rain rate trend in South Iraq, 1999-2012

Figure 5: Rain rate trend in South Iraq, 1999-2012

Figure 6: Rain rate trend in North West Saudi Arabia, 1999-2012

Figure 6: Rain rate trend in North West Saudi Arabia, 1999-2012

Figure 7: Rain rate trend in South Somalia, 1999-2012

Figure 7: Rain rate trend in South Somalia, 1999-2012

Figure 8: Rain rate trend in North Somalia, 1999-2012

Figure 8: Rain rate trend in North Somalia, 1999-2012

Figure 9: Rain rate trend in Middle Sudan, 1999-2012

Figure 9: Rain rate trend in Middle Sudan, 1999-2012

Figure 10: Rain rate trend in Sudd region-Sudan, 1999-2012

Figure 10: Rain rate trend in Sudd region-Sudan, 1999-2012

Figure 11: Rain rate trend in Lebanon and Jordan, 1999-2012

Figure 11: Rain rate trend in Lebanon and Jordan, 1999-2012

Conclusion

33This paper presents a methodology for the characterization and analysis of satellite-images time series. The significance of this research lies in the integration of a GIS framework that can enable geographic information integration in a smooth and flexible way for the assessment of land degradation using remote sensing large databases. This integration considers spatial and temporal information included in an NDVI time series data. However, the usage of the NDVI as the only factor for this study presents some limitations: the vegetation changes can be caused by different variables other than the NDVI.

34All results and outputs are published in a biennial bulletin (CNRS, ACSAD and GIZ, 2015) for land degradation monitoring and assessment in the Arab world, produced under the joint collaboration between the Lebanese CNRS, the ACSAD, and sponsored by German International Cooperation (GIZ).

35The objective of this study was to investigate the general applicability of the time-series (SPOT) NDVI datasets for the assessment of land degradation in the MENA area. NDVI datasets were found to have sufficient spatial, spectral, and temporal resolutions to detect unique multi-temporal signatures, allowing for seasonal response tracking. As for the MENA region, many Arab countries have been aware of land degradation issues and tried to resolve them, while others have been more pressing problems to be concerned with. Nonetheless, in general, we demonstrated that the MENA area has a growing degraded vegetation area, which can exacerbate directly problems like agriculture, human migration and other’s indirect concerns touching the socioeconomic status of these countries and their citizens.

36Most Arab countries show a higher degraded area than a developed one. Almost 553 000 km2 of their total surface presents either a negative change or a hotspot. This reflects the critical and dangerous situation that this region is facing.

Haut de page

Bibliographie

Afifi, A., A. Gad, and A. Refat, 2010, "Use of GIS and remote sensing for environmental sensitivity assessment of north coastal part, Egypt." Journal of American Science, Vol.6, No.11, 632-646.

Aldakheel, Yousef Y.,2011, "Assessing NDVI spatial pattern as related to irrigation and soil salinity management in Al-Hassa Oasis, Saudi Arabia." Journal of the indian Society of Remote Sensing, Vol.39, No.2, 171-180.

Alves, D. Borini, F. Pérez-Cabello, and M. Rodrigues Mimbrero, 2015, "Land-use and land-cover dynamics monitored by NDVI multitemporal analysis in a selected southern Amazonian area (Brazil) for the last three decades." The International Archives of Photogrammetry, Remote Sensing and Spatial Information Sciences, Vol.40, No.7, 329.

Ardö, Jonas, 2015, "Comparison between remote sensing and a dynamic vegetation model for estimating terrestrial primary production of Africa." Carbon balance and management, Vol.10, No.1, 8.

Artan, G., H. Gadain, F. Muthusi, and P. Muchiri, 2007, Improving flood forecasting and early warning in Somalia, Technical report 96 pages. http://www.faoswalim.org/resources/site_files/W-10%20Improving%20Flood%20Forecasting%20and%20Early%20Warning%20in%20Somalia.pdf

Bai, Zhanguo G., David L. Dent, Lennart Olsson, and Michael E. Schaepman, 2003, "Proxy global assessment of land degradation." Soil use and management, Vol.24, No.3, 223-234.

Balanche, Fabrice, 2013, "La modernisation des systèmes d'irrigation dans le Nord‑Est syrien: la bureaucratie au cœur de la relation eau et pouvoir." Méditerranée, Vol.119, No.2, 59-72.

Becker, Richard H.,2014, "The Stalled Recovery of the Iraqi Marshes." Remote Sensing, Vol.6, No.2, 1260-1274.

Berkat O., Tazi M., 2004, "Country Pasture/Forage Resource Profiles", FAO, http://www.fao.org/ag/Agp/agpc/doc/counprof/PDF%20files/Morocco-English.pdf

Brown, Molly E., Jorge E. Pinzon, and Compton J. Tucker, 2004, "New vegetation index data set available to monitor global change." Transactions American Geophysical Union, Vol.85, No.52, 565-569.

Brown, Molly E., Jorge E. Pinzón, Kamel Didan, Jeffrey T. Morisette, and Compton J. Tucker, 2006, "Evaluation of the consistency of long-term NDVI time series derived from AVHRR, SPOT-vegetation, SeaWiFS, MODIS, and Landsat ETM+ sensors." Geoscience and Remote Sensing, IEEE Transactions, Vol.44, No.7, 1787-1793.

Chen, Xiao, Bo Qiu, Xuhua Cheng, Yiquan Qi, and Yan Du, 2015, "Intra-seasonal variability of Pacific-origin sea level anomalies around the Philippine Archipelago." Journal of Oceanography, Vol.71, No.3, 239-249.

CNRS (Centre National de la Recherche Scientifique-Liban), ACSAD (Arab Center for the Studies of Arid Zones and Dry Lands) and GIZ (German International Cooperation), 2015, “Desertification Bulletin 2015”, 16 pages.

Daccache, A., J. S. Ciurana, JA Rodriguez Diaz, and J. W. Knox, 2014, "Water and energy footprint of irrigated agriculture in the Mediterranean region." Environmental Research Letters, Vol.9, No.12, 1-12.

Donohue, R. J., I. H. Hume, M. L. Roderick, T. R. McVicar, Jason Beringer, L. B. Hutley, J. C. Gallant et al., 2014, "Evaluation of the remote-sensing-based DIFFUSE model for estimating photosynthesis of vegetation." Remote Sensing of Environment, Vol.155, 349-365.

Du, Jinyang, John S. Kimball, and Lucas A. Jones, 2016, "Passive Microwave Remote Sensing of Soil Moisture Based on Dynamic Vegetation Scattering Properties for AMSR-E." Geoscience and Remote Sensing, IEEE Transactions, Vol.54, No.1, 597-608.

Elmzoughi, Amor, Mehriz Zribi, Riadh Abdelfattah, and Ziad Belhadj, 2008, "Using long term NOAA-AVHRR NDVI time-series data for desertification monitoring in South Tunisia." Information and Communication Technologies: From Theory to Applications, ICTTA 2008. 3rd International Conference, 1-6.

Evans, Jason, and Roland Geerken, 2004, "Discrimination between climate and human-induced dryland degradation." Journal of Arid Environments, Vol.57, No.4, 535-554.

Fensholt, Rasmus, Kjeld Rasmussen, Per Kaspersen, Silvia Huber, Stephanie Horion, and Else Swinnen, 2013, "Assessing land degradation/recovery in the African Sahel from long-term earth observation based primary productivity and precipitation relationships." Remote Sensing, Vol.5, No.2, 664-686.

Fensholt, Rasmus, and Simon R. Proud, 2012, "Evaluation of earth observation based global long term vegetation trends—Comparing GIMMS and MODIS global NDVI time series." Remote sensing of Environment, Vol.119, 131-147.

Friedl, Mark A., Douglas K. McIver, John CF Hodges, X. Y. Zhang, D. Muchoney, Alan H. Strahler, Curtis E. Woodcock et al., 2002, "Global land cover mapping from MODIS: algorithms and early results." Remote Sensing of Environment 83, No.1, 287-302.

Funk, Chris C., and Molly E. Brown, 2006, "Intra-seasonal NDVI change projections in semi-arid Africa." Remote Sensing of Environment, Vol.101, No.2, 249-256.

Geerken, Roland, and Mohammad Ilaiwi, 2004, "Assessment of rangeland degradation and development of a strategy for rehabilitation." Remote Sensing of Environment, Vol.90, No.4, 490-504.

Gillies, Robert R., and Toby N. Carlson, 1995, "Thermal remote sensing of surface soil water content with partial vegetation cover for incorporation into climate models." Journal of Applied Meteorology, Vol.34, No.4, 745-756.

Haboudane, Driss, John R. Miller, Elizabeth Pattey, Pablo J. Zarco-Tejada, and Ian B. Strachan, 2004, "Hyperspectral vegetation indices and novel algorithms for predicting green LAI of crop canopies: Modeling and validation in the context of precision agriculture." Remote sensing of environment, Vol.90, No.3, 337-352.

Huete, A. R., and C. J. Tucker, 1991, "Investigation of soil influences in AVHRR red and near-infrared vegetation index imagery." International Journal of Remote Sensing, Vol.12, No.6, 1223-1242.

IPCC. Technical paper. 2008. Accessed on 6 July 2013. URL: http://www.ipcc.ch/

Jabbar, Mushtak T., and Jing-xuan Zhou, 2013, "Environmental degradation assessment in arid areas: a case study from Basra Province, southern Iraq." Environmental earth sciences, Vol.70, No.5, 2203-2214.

Jamali, Sadegh, Jonathan Seaquist, Lars Eklundh, and Jonas Ardö, 2012, "Comparing parametric and non-parametric approaches for estimating trends in multi-year NDVI." In 1st EARSeL Workshop on Temporal Analysis of Satellite Images.

Julien, Yves, and José Sobrino, 2007, "Changes in the global vegetal cover through a phenological analysis of GIMMS data." Analysis of Multi-temporal Remote Sensing Images, 1-5.

Khader, M., Mederbal, K., and Chouieb, M, 2014, "Suivi de la dégradation de la végétation steppique à l’aide de la télédétection: cas des parcours steppiques région de Djelfa (Algérie)." University of Biskra Repository.

Kouadio, Louis, Nathaniel K. Newlands, Andrew Davidson, Yinsuo Zhang, and Aston Chipanshi, 2014, "Assessing the performance of MODIS NDVI and EVI for seasonal crop yield forecasting at the ecodistrict scale." Remote Sensing, Vol.6, No.10, 10193-10214.

Li, Zhe, Ted Huffman, Brian McConkey, and Lawrence Townley-Smith, 2013, "Monitoring and modeling spatial and temporal patterns of grassland dynamics using time-series MODIS NDVI with climate and stocking data." Remote Sensing of Environment, Vol.138, 232-244.

Los, S. O., 2013, "Analysis of trends in fused AVHRR and MODIS NDVI data for 1982–2006: Indication for a CO2 fertilization effect in global vegetation." Global Biogeochemical Cycles, Vol.27, No.2, 318-330.

Luo, Xiangzhong, Xiaoqiu Chen, Lin Xu, Ranga Myneni, and Zaichun Zhu, 2013, "Assessing performance of NDVI and NDVI3g in monitoring leaf unfolding dates of the deciduous broadleaf forest in Northern China." Remote Sensing, Vol.5, No.2, 845-861.

Maselli, F., and Rembold, F. Forecasting, Yield, 2001. "Analysis of GAC NDVI data for cropland identification and yield forecasting in Mediterranean African countries." Photogrammetric Engineering & Remote Sensing, Vol.67, No. 5, 593-602.

Moncrieff, G. R., S. Scheiter, J. A. Slingsby, and S. I. Higgins, 2015, "Understanding global change impacts on South African biomes using Dynamic Vegetation Models." South African Journal of Botany, Vol.101,16-23.

Neeti, Neeti, and J. Ronald Eastman, 2011, "A contextual mann‐kendall approach for the assessment of trend significance in image time series." Transactions in GIS, Vol.15, No.5, 599-611.

Nielsen, Thomas Theis, and H. K. Adriansen, 2005, "Government policies and land degradation in the Middle East." Land Degradation & Development, Vol.16, No.2, 151-161.

Othman, Arsalan A., Younus I. Al-Saady, Ahmed K. Al-Khafaji, and Richard Gloaguen, 2014, "Environmental change detection in the central part of Iraq using remote sensing data and GIS." Arabian Journal of Geosciences, Vol.7, No.3, 1017-1028.

Pouliot, Darren, Rasim Latifovic, Natalie Zabcic, Luc Guindon, and Ian Olthof, 2014, "Development and assessment of a 250m spatial resolution MODIS annual land cover time series (2000–2011) for the forest region of Canada derived from change-based updating." Remote Sensing of Environment, Vol.140, 731-743.

Rasmussen, Kjeld, Thomas Theis Nielsen, Cheikh Mbow, and Andrew Wardell, 2006, Land degradation in the Sahel: An apparent scientific contradiction, 17th Danish Sahel Workshop.

Rasmussen, Michael S., 1997, "Operational yield forecast using AVHRR NDVI data: reduction of environmental and inter-annual variability." International Journal of Remote Sensing, Vol.18, No.5, 1059-1077.

Ronald Eastman, J., Florencia Sangermano, Bardan Ghimire, Honglei Zhu, Hao Chen, Neeti Neeti, Yongming Cai, Elia A. Machado, and Stefano C. Crema, 2009, "Seasonal trend analysis of image time series." International Journal of Remote Sensing, Vol.30, No.10, 2721-2726.

Seiler, R. A., F. Kogan, and Guo Wei, 2000, "Monitoring weather impact and crop yield from NOAA AVHRR data in Argentina." Advances in Space Research, Vol.26, No.7, 1177-1185.

Seiler, R. A., F. Kogan, Guo Wei, and M. Vinocur, 2007, "Seasonal and interannual responses of the vegetation and production of crops in Cordoba–Argentina assessed by AVHRR derived vegetation indices." Advances in Space Research, Vol. 39, No.1, 88-94.

Shao, Yang, James B. Campbell, Gregory N. Taff, and Baojuan Zheng, 2015, "An analysis of cropland mask choice and ancillary data for annual corn yield forecasting using MODIS data." International Journal of Applied Earth Observation and Geoinformation, Vol,38, 78-87.

Stockli, Reto, and Pier Luigi Vidale, 2004, "European plant phenology and climate as seen in a 20-year AVHRR land-surface parameter dataset." International Journal of Remote Sensing, Vol.25, No.17, 3303-3330.

Tarnavsky, Elena, Sebastien Garrigues, and Molly E. Brown, 2008, "Multiscale geostatistical analysis of AVHRR, SPOT-VGT, and MODIS global NDVI products." Remote Sensing of Environment, Vol.112, No.2, 535-549.

Terink, Wilco, Walter Willem Immerzeel, and Peter Droogers, 2013, "Climate change projections of precipitation and reference evapotranspiration for the Middle East and Northern Africa until 2050." International journal of climatology, Vol.33, No.14, 3055-3072.

Tian, Feng, Yunjia Wang, Rasmus Fensholt, Kun Wang, Li Zhang, and Yi Huang, 2013, "Mapping and evaluation of NDVI trends from synthetic time series obtained by blending Landsat and MODIS data around a coalfield on the Loess Plateau." Remote Sensing, Vol.5, No. 9, 4255-4279.

Tian, Feng, Rasmus Fensholt, Jan Verbesselt, Kenneth Grogan, Stephanie Horion, and Yunjia Wang, 2015, "Evaluating temporal consistency of long-term global NDVI datasets for trend analysis." Remote Sensing of Environment, Vol.163, 326-340.

Townshend, John RG, 1994, "Global data sets for land applications from the Advanced Very High Resolution Radiometer: an introduction." International Journal of Remote Sensing, Vol.15, No.17, 3319-3332.

Trigo, Ricardo M., Célia M. Gouveia, and David Barriopedro, 2010, "The intense 2007–2009 drought in the Fertile Crescent: Impacts and associated atmospheric circulation." Agricultural and Forest Meteorology, Vol.150, No. 9, 1245-1257.

Tucker, Compton J., Jorge E. Pinzon, Molly E. Brown, Daniel A. Slayback, Edwin W. Pak, Robert Mahoney, Eric F. Vermote, and Nazmi El Saleous, 2005, "An extended AVHRR 8‐km NDVI dataset compatible with MODIS and SPOT vegetation NDVI data." International Journal of Remote Sensing, Vol.26, No.20, 4485-4498.

Udelhoven, Th, and J. Hill, 2009, "Change detection in Syria’s rangelands using long-term AVHRR data (1982–2004)." Recent advances in remote sensing and geoinformation processing for land degradation assessment, 117-132.

Walker, J. J., K. M. de Beurs, and R. H. Wynne, 2014, "Dryland vegetation phenology across an elevation gradient in Arizona, USA, investigated with fused MODIS and Landsat data." Remote Sensing of Environment, Vol.144, 85-97.

White, J. C., M. A. Wulder, G. W. Hobart, J. E. Luther, T. Hermosilla, P. Griffiths, N. C. Coops et al., 2014, "Pixel-based image compositing for large-area dense time series applications and science." Canadian Journal of Remote Sensing, Vol.40, No.3, 192-212.

Zhou, Liming, Compton J. Tucker, Robert K. Kaufmann, Daniel Slayback, Nikolay V. Shabanov, and Ranga B. Myneni, 2001, "Variations in northern vegetation activity inferred from satellite data of vegetation index during 1981 to 1999." Journal of Geophysical Research: Atmospheres (1984–2012), Vol.106, No.D17, 20069-20083.

Haut de page

Table des illustrations

Titre Figure 1: Territories of the Arab countries
URL http://journals.openedition.org/cybergeo/docannexe/image/27620/img-1.png
Fichier image/png, 2,4M
Titre Figure 2: Areas with vegetation changes based on the SPOT Vegetation NDVI dataset, 1999-2012
URL http://journals.openedition.org/cybergeo/docannexe/image/27620/img-2.png
Fichier image/png, 1,2M
Titre Figure 3: Rain rate trend in North Morocco, 1999-2012
URL http://journals.openedition.org/cybergeo/docannexe/image/27620/img-3.png
Fichier image/png, 44k
Titre Figure 4: Rain rate trend in North Algeria and Tunisia, 1999-2012
URL http://journals.openedition.org/cybergeo/docannexe/image/27620/img-4.png
Fichier image/png, 48k
Titre Figure 5: Rain rate trend in South Iraq, 1999-2012
URL http://journals.openedition.org/cybergeo/docannexe/image/27620/img-5.png
Fichier image/png, 38k
Titre Figure 6: Rain rate trend in North West Saudi Arabia, 1999-2012
URL http://journals.openedition.org/cybergeo/docannexe/image/27620/img-6.png
Fichier image/png, 39k
Titre Figure 7: Rain rate trend in South Somalia, 1999-2012
URL http://journals.openedition.org/cybergeo/docannexe/image/27620/img-7.png
Fichier image/png, 43k
Titre Figure 8: Rain rate trend in North Somalia, 1999-2012
URL http://journals.openedition.org/cybergeo/docannexe/image/27620/img-8.png
Fichier image/png, 46k
Titre Figure 9: Rain rate trend in Middle Sudan, 1999-2012
URL http://journals.openedition.org/cybergeo/docannexe/image/27620/img-9.png
Fichier image/png, 36k
Titre Figure 10: Rain rate trend in Sudd region-Sudan, 1999-2012
URL http://journals.openedition.org/cybergeo/docannexe/image/27620/img-10.png
Fichier image/png, 44k
Titre Figure 11: Rain rate trend in Lebanon and Jordan, 1999-2012
URL http://journals.openedition.org/cybergeo/docannexe/image/27620/img-11.png
Fichier image/png, 38k
Haut de page

Pour citer cet article

Référence électronique

Ghaleb Faour, Mario Mhawej et Abbas Fayad, « Detecting Changes in Vegetation Trends in the Middle East and North Africa (MENA) Region Using SPOT Vegetation », Cybergeo: European Journal of Geography [En ligne], Cartographie, Imagerie, SIG, document 779, mis en ligne le 14 avril 2016, consulté le 19 mars 2024. URL : http://journals.openedition.org/cybergeo/27620 ; DOI : https://doi.org/10.4000/cybergeo.27620

Haut de page

Auteurs

Ghaleb Faour

National Center for Remote Sensing, National Council for Scientific Research (CNRS), Beirut, Lebanon
gfaour@cnrs.edu.lb

Mario Mhawej

National Center for Remote Sensing, National Council for Scientific Research (CNRS), Beirut, Lebanon
Author to whom correspondence should be addressed
mario.mhawej@gmail.com
Tel.: +961-4-409-845; Fax: +961-4-409-847.

Abbas Fayad

National Center for Remote Sensing, National Council for Scientific Research (CNRS), Beirut, Lebanon
abbasfayad@yahoo.com

Haut de page

Droits d’auteur

CC-BY-4.0

Le texte seul est utilisable sous licence CC BY 4.0. Les autres éléments (illustrations, fichiers annexes importés) sont « Tous droits réservés », sauf mention contraire.

Haut de page
Rechercher dans OpenEdition Search

Vous allez être redirigé vers OpenEdition Search