Application of Remote sensing and GIS on studying the impact of « Jasmin revolution » on Mellegue catchment

doi:10.21203/rs.3.rs-2564181/v1

“Jasmin revolution” is a Tunisian revolt occurred in 2011 referring to the first uprising that unleashed similar events in Middle East and North-Africa called “Arab Spring”. This revolution has led to series of political actions where the government has passed through consecutives bumps that caused socio-political insecurity and environmental damage. The watershed of Mellegue is a cross-border basin belonging to Algeria and Tunisia. It is a vast catchment occupying more than 10500 km². This paper aimed to study the post-revolution impacts on the watershed Mellegue using geospatial data, remote sensing and GIS. It highlights that the resources were significantly consumed. Remote sensing analysis showed that forests were reduced due to anthropogenic activities. Water quantification revealed an over-consumption of these resources where the consumption rate was intensified after 2011 revolution. Population census using geospatial data and modelling approach exhibits a total increase in demography of the study area but also exposes an antagonist behaviour in Algerian territory compared to the Tunisian lands of the catchment. The Algerian demography are in consistent rise unlike Tunisian population that showed a stable to even decline in population growth.

2011 revolution

watershed of Mellegue

remote sensing

GIS

resources

Arab Spring was a series of pro-democracy uprisings that took place in the Middle East and North Africa beginning in 2010 and continuing in 2011. It was challenging some of the region’s entrenched authoritarian regimes. The wave began with “Jasmin revolutionary” protesting the Tunisian political regimes which led to the former president Zine El-Abidine Ben Ali that ruled for more than 20 years to abdicate this position and run away to Saudi Arabia. These events have inspired other Arab countries like Egypt, Yemen, Syria, Libya and Bahrain to express their political and economic grievances. Tunisian have lived since a series of democratic elections with multiple presidents through the last decade (2011–2022). Even though not every country saw success in protesting where they were often met with violent crackdowns by their countries’ security forces (Fraihat 2016).

The impacts of the events in these nations were seen first in northern Africa and continue in the Middle East. However, the political and social impact of these popular uprisings remains significant today, due to the series of large-scale political and social instabilities.

After a decade of Tunisian Jasmine revolutionary, the government have passed through a set of socio-political and economic crisis. Even though democracy and cultural freedom was increased but the repercussions were seen in different aspects including anthropogenic and terrorist actions that affected the environment, the ecology and the resources. In that context, the watershed of Mellegue define a transboundary catchment belonging to Algeria and Tunisia. It starts in Algeria with more than 60% of the area is belonging to that country where the rest (including the outlet) is situated in Tunisia. The catchment is very vast exceedingly more than one million hectare of surface with a rich fauna and flora variety. The degradation of catchment is being noticed lately with a strong decrease in resources and serious decline in ecosystem (Mlayah et al. 2017; Mlayah et al. 2013).

Human has always showed the capacity to adapt to all of the Earth’s environment. Moreover, the human habitat for each of these environments remains strongly clustered (Martine and Schensul 2013). This fact will certainly affect the earth’s landscapes due to the anthropogenic impact where mapping it is often challenging. On the other hand, the mapping of demographic population within a certain geophysical parameter space can be effective in quantifying the breadth and the clustering nature of human adaptation across environments (Small and Sousa 2016).

The amount of land use/land cover intensity changes can be influenced due to many factors especially anthropogenic (Bucała 2014; Mahmoud and Gan 2018). As a consequence, is it always difficult to measure the rate of landscapes alteration. Remote sensing and Geographic Information System (GIS) offers a variety of tools to monitor and study different types of anthropogenic changes and gives a precise information about the rate of disturbance (Small and Sousa 2016). When combined to a certain expertise and local level knowledge, these tools can provide features of landscapes regionally or globally, tolerating insights for more research perspectives and in depth-study. Anthropologists concentrate essentially on fieldwork and cultural knowledge that provides an extra dimension to the final analysis. Therefore, anthropocentric viewpoints of environmental degradation apply human outcomes of change for measuring criticality. This infers to the fact that most critical regions were always threatened by the long-term sustainability of the human environment bond. In that matter, remote sensing can offer accurate and continuous information about landscapes under spatial-temporal study. GIS can rather be important in analysing, storing and extract modelling data parameters of remote sensing data (Liping et al. 2018; Rawat and Kumar 2015). Moreover, GIS is implemented to concept spatial models locally as well globally which will be helpful subsequently for decision-making or updating environmental policies.

The present study was aimed to study the impacts of the socio-political troubles of 2011 Tunisian “Jasmin revolution”, that initiated “Arab Spring”, on the environment and ecosystem of the Mellegue watershed. The study was based on remote sensing analysis using muti-temporal image series and geospatial data to compare earlier to post catchment resources. The temporal range of the study extend from 2002 to 2021 as way to put 2011 data as midpoint that fairly split the hole interval date to pre and post revolution and to fetch for any potential resource unbalance compared to the normal trend.

Study Area

Mellegue catchment is limited between 35°12’20.7”, 36°25’50.43” North, 7°11’30.98”, 8°55’8° East. It is a large cross-border catchment occupying 10500 km² of surface and 1054 km of perimeter. More than sixty percent of the surface (6255 km²) is belonging to Algerian government. The remaining surface as well as the outlet are located in Tunisia. The principal river of the watershed is known by the so-called Oued Mellegue. It starts in the mounts of Tebessa, flows through the famous Algerian Aures massif (Belloula and Dridi 2015). It continues the path in Tunisia, through the long mountain range of Tunisian Atlas which separates north from Tunisian midland, until finally join the river of Mejerdah in Jendouba at 140 m height above the sea level. The overall direction follows the southwestern-northeastern direction of the Algerian-Tunisian Atlas (Mlayah et al. 2009). The overall shape of the catchment is elongated, explicitly large in the middle, with an overall NE-SW orientation. The slope of the catchment generally varies from 0 to 25%. However, the basin holds low to moderate relief where almost 70% of the surface is characterized with low slope. Only 3% of the total surface hold a strong slope.

As for the geology of the watershed, it contains a variety of ages where identified outcrops dating extend from Trias to Quaternary with the exception of Jurassic layers. The basin is also known of many evaporitic salt extrusions due to the action of border faults that extrudes the Trias layers of evaporate and silicates. The structural study of the area reveals a complex and assortment geologic shapes such as diapir area that cause a set of NE-SW oriented folds located generally on the Algerian-Tunisian border. Other famous structures also exist like Tunisian Atlas zone or the Neogene folds (Rouvier 1977).

The catchment is occupied by a sparse vegetation cover. The mountains contain a diversified vegetation, the foothills are undeveloped while the plains are almost cultivated. The densest forests rarely occur, especially downstream of the basin. On the other hand, bare areas and cultivated soils are the most apparent, covering a large part of the basin. The watershed also holds some vegetations from salty wetlands and sabkhas which reflect poor flow and drainage which refers to a strong accumulation of salty water (J.A.Rodier et al. 1981).

The general climate of the basin is arid and semi-arid, locally subhumid at the extreme north. It belongs to the Mediterranean climate which is generally characterized by moderate, humid winters and hot, dry summers. Rainfalls are highly variable depending on the region, extending from 400 mm/year to more than 700 mm/year. The maximum temperatures reach an average of 39°C (plus or minus 3°C) while the minimum temperatures vary from − 2 to 8°C. The lowest temperatures are mostly recorded in the west and north of the basin where it often snows (Belloula 2017).

This Mediterranean climate is tightly blocked between a cooler temperate domain in the North and the desert in the South except at the very edge of the sea where we do not observe a true maritime climate, but rather a contrasting climate reflecting a continental influence rather than oceanic, reflecting above all their double belonging to the fringes of the temperate and the tropical zone.

To study the resources of Mellegue catchment, various materials were applied. Main adopted tools were cited in the following sections. In addition, other ancillary data were required like statistic data from local official headquarters, areal imageries and other supplementary materials (digital elevation model, topographic and geologic maps, etc.).

Geospatial data and tools for population census in Mellegue catchment

The general population census (in space and time) is an essential step which constitutes the first source of reliable, detailed and precise data on the socio-economic characteristics of a given region (developments, housing, living standards). Unfortunately, traditional data are sometimes outdated and takes a lot of time and resources to input. It was in 2013 that a program called WorldPop (worldpop 2013) was created and launched, which dispose of a large number of data and new methodologies for mapping and calculating population density. The population mapping is based on Random Forest model (Stevens et al. 2015) which provides population densities through semi-automatic algorithms operated on Landsat satellite imagery (ETM+). The adopted census model is based on a free estimation method called unconstrained estimation modelling approach (Gaughan et al. 2016; Sorichetta et al. 2015; Stevens et al. 2015) where it was developed, refined and tested for five years. Population density is estimated per grid cell. The dataset is available in raster (GeoTIFF) and text (ASCII XYZ) format with 30 arc seconds resolution (about 1 km at the equator). The unit of measurement is expressed in number of people per square kilometre. Computed census results of the populations are correlated, confirmed and corrected with the corresponding official population estimates of the United Nations, which were initially prepared and processed by the Population Division by the specialized United Nations Secretariat Department of Economic and Social Affairs.

Since the data is available by country. The methodology applied to estimate the population density of the Mellegue watershed by collecting population data from Algeria and Tunisia. It will then be necessary to merge these data by focusing only on the demographic population that exists within the basin. This task was carried out by ArcGIS where each year is treated separately. All data are exported in text format.

Geospatial data and tools for forest resource analysis

Normalized Difference Vegetation Index (NDVI)

NDVI plays a very important role in remote sensing analysis. The easy and simple use made it important for quantitative and qualitative vegetation measurements (Bhandari et al. 2012). It is often used globally to map highly productive areas, as well as to discover weakened soils that require special treatment (Singh et al. 2004). It is also used to monitor drought, forecast agricultural yield, predict fire zones and map desertification trends (de Carvalho et al. 2014; Illera et al. 1996; Krtalic et al. 2019). Agricultural applications, such as crop monitoring, incorporate NDVI to aid in crop survey and to provide accuracy for fertilizer application or irrigation, among other treatment activities, at growth stages (Singh et al. 2004). However, NDVI is specifically chosen for general vegetation monitoring for the reason that it compensates many factor alterations (lighting conditions, surface slope, exposure, and other external factors). Crop monitoring from NDVI tracks seasonal changes in fields based on 7 years historical field productivity (EOS Data Analytics 2023). This index is useful in monitoring patterns of crop rotation as well as current vegetation rates. Using user-friendly graphs, the NDVI application can visualize different types of data including vegetation indices, growth stages, historical and present weather parameters (rainfalls rate, temperature) among others.

NDVI is a measure of plant health revealing how its capacity to reflect light at certain frequencies (some waves are absorbed and some are reflected). Chlorophyll (an indicator of health) intensely absorbs visible light. Oppositely, the cellular shape of leaves deeply reflects near-infrared. When the plant is dehydrated or sick (afflicted with diseases), the spongy layer weakens causing the plant absorbing more near infrared light, rather than reflecting it. Thus, observing how near infrared (NIR) changes compared to red light provides an accurate indication of the presence of chlorophyll, which correlates with plant health. Consequently, the calculation of NDVI is based on the NIR and red waves using the Hill and Donald (2003) equation:

$$NDVI= \frac{NIR-RED}{NIR+RED}$$

NIR: near infrared

According to this formula, the vegetation density (NDVI) in a specific coordinate of the image equals the intensities’ difference of the reflected lights in the red and the near infrared divided by the sum of these intensities. This index has a range of values between − 1 and 1, representing essentially vegetation, where negative values refer to unvegetated areas (clouds, water and snow) while values on the average of zero define bare grounds. Moderate values (0.2 to 0.3) refer to shrubs and grasslands, while high values (0.6 to 0.8) refer to dense vegetation. Crop monitoring successfully uses this NDVI interval to show users which area have dense, moderate, or sparse vegetation at any given time.

Fire Information for Resource Management System (FIRMS)

NASA-FIRMS constitutes a database collecting fires, dedicated for resource management. This database is based on satellite observations from VIIRS and MODIS instruments to detect active fires and thermal anomalies. The generated results are provided in near real time (3 hours delay after the satellite observation). It can be transmitted in different ways, either by email, analysis ready data, web services or even online maps (NASA 2022).

Moderate Resolution Imaging Spectroradiometer (MODIS)

MODIS is an embedded instrument on board of Terra (initially called EOS AM-1) and Aqua (initially named EOS PM-1) satellites. Terra orbits around the Earth from North to South across the equator in the morning, while Aqua orbits from South to North above the equator in the afternoon. This makes Terra and Aqua MODIS observe the entire surface of the Earth every 1–2 days, gathering data information in 36 spectral bands or wavelength groups.

Fire detection is performed using a complex algorithm that perceives the strong emission of mid-infrared radiation from fires (Giglio et al. 2003). The algorithm examines each pixel in the MODIS band and finally assigns each one a specific class among the following: missing data, unknown, cloud, no fires, water, or fires. Pixels with no valid data will be immediately classified as missing data and excluded from further analysis. Cloud and water pixels are identified using cloud and water masks, and will be assigned by their class names respectively. Processing continues on the remaining pixels of the light spots.

A preliminary classification is used to eliminate obvious pixels that confirms the lack of fire. Regarding the remaining potential fire pixels, an approach is made based on the neighbouring pixels to estimate the radiometric signal of the possible fire pixel under normal conditions (absence of fire). Valid neighbouring pixels within a centred window on the probable fire pixel are spotted and correlated to approximate a background value. If the background rate estimation is effective, a series of background threshold tests are applied to perform relative fire detection. This is generally based on the distinguishing feature of an active fire in which 4µm range of brightness temperature and the brightness temperature difference of 4 and 11µm deviate significantly from the background estimated value without fire. The relative thresholds are adjusted according to the natural variability of the background noise. Additional specific tests are employed subsequently to eradicate false detections made by sun reflectance, desert boundaries, as well as water mask errors. Non-rejected pixels during the application of these tests are assigned as fire class. Invalid pixels either from unsuccessful background characterization or those with an insufficient valid pixel’s number, are assigned unknown.

Tolerable MODIS cover range of detect active is approximately 1000 m². Under very good viewing conditions (i.e., close to nadir, absence (few) of smoke, or in case of homogeneous land surface), small flaming fires of 100 m² can be detected. Under perfect (and extremely rare) viewing conditions, even smaller flaming fires of around 50 m² can be identified.

Visible Infrared Imaging Radiometer Suite (VIIRS)

The NASA Suomi-NPP VIIRS Active Fire product suite is built on MODIS algorithm (MOD14/MYD14) to detection of fires and thermal anomalies (Giglio et al. 2003). The product is systematically mapping the global activity of fires at 12 hours intervals. Two active VIIRS fire products are created independently based on the 750 and 375 m available data resolution. The VIIRS-750m product define a replica of the Collection 6 of MODIS algorithm, which has been adapted to the VIIRS M5 to M16 chains. The product includes the detection and active distinction of fires using fire radiative power (FRP) on land and water (gas flames). In addition, this product is very useful, as it provides continuous data, especially when it combines with the Earth Observation System (EOS) MODIS-1km product.

The 375m alternative product uses the higher spatial resolution VIIRS I1-I5 channels, supplemented by the M13 channel, to detect and depict sub-pixel active fires. It resembles the MOD14/MYD14 fire products, with the addition of specific tests and metadata to account for the unique characteristics of the VIIRS-375m tapes. Compared to other resolution (~ 1km) satellite products for fire detection, the VIIRS-375m product provides better response when detecting small fires, as well as better perimeter mapping of large fires. Therefore, the data is well suited for use in support of fire management and near real-time wildfire warning systems, as well as other scientific applications that require better reliable mapping fires.

Geospatial data and tools for water catchment quantification

Normalized difference water index (NDWI)

The NDWI factor was firstly introduced in 1996 by Gao (1996). It exhibits the moisture content of plants and soil. NDWI is widely applied, starting from agricultural monitoring like crop irrigation and pasture management, to forest monitoring by fire risk management as well as fuel moisture monitoring (live fuel moisture) which is essential for climate change studies (Gu et al. 2008; Gu et al. 2007; Qiao et al. 2012).

Vegetation is under severe water stress during periods of drought due to climate change and global warming. In fact, entire crops can be damaged if affected areas are not identified. For that reason, early detection of water stress can avoid many severe impacts on crops. Remote sensing observation of land combined with the NDWI can control irrigation in real time, significantly improving agriculture, especially in areas where it is difficult to meet water needs (Deshayes et al. 2020).

$$NDWI= \frac{NIR-SWIR}{NIR+SWIR}$$

With: NIR: near infrared (0.841–0.876 nm)

SWIR: Short Wave Infrared (1.628–1.652 nm)

Instead of using NDVI's red range, whose reflective intensity is determined by the presence of chlorophyll, NDWI's calculation relies on short-wave near-infrared (SWIR) in which strong light absorption by water occurs. The use of near infrared (NIR) in both NDVI and NDWI indexes is due to the fact that water does not absorb this kind of electromagnetic spectrum, which makes NDWI index resistant to atmospheric effects, distinguishing it from NDVI. The NDWI index is using in the observation of forests. It is a good indicator of the ecological condition of forests especially when critical anthropogenic load is reached.

The NDWI index is preferred in the monitoring of fire risks by measuring the plant cover humidity. Higher NDWI values indicate sufficient moisture, while low values point to water stress. NDWI results can be presented in maps and graphs, providing valuable information on the spatial spreading of vegetation water stress and its temporal evolution over longer periods. The NDWI product is dimensionless and goes from − 1 to + 1, depending on the water content, as well as the type of vegetation. high NDWI values indicate high plant water content (high plant fraction) covering while low NDWI amounts correspond to low water content.

Gravity Recovery and Climate Experiment (GRACE/GRACE-FO)

GRACE is a partnership space mission between NASA and Geosciences German Research Centre (GFZ). It was launched to track the water height of the earth to monitor changes in groundwater storage, lakes and rivers water amount, soil moisture, ice covers and ocean sea level surveillance. These discoveries offer a unique insight of the earth's climate, providing significant benefits for science and society.

The GRACE concept is based on comparing measured gravity according to the distance between the twin launched satellites. Any observed mass gravity variations are mostly due to the changing thickness of very thin layers water, located near the surface of the Earth (might reach several kilometres thick). When an increase in gravity is noticed (greater concentration of mass) in front of the pair, the front satellite accelerated causing an increase of the gap between the pair. On the opposite, when the gravity between the pairs increase, their distance will decrease. The total reverse will happen when there was a decrease in gravity in front of, or between, the pair satellites. The satellites are separated by an exact distance of 220 km where it can detect changes in relative speed of less than one micrometre per second. These measurements, in conjunction with other data and models, have provided observations of terrestrial water storing changes, including ice mass variation, ocean floor pressure vicissitudes, and sea level fluctuation (Landerer and Cooley 2019).

GRACE-FO defines the successor of the initial GRACE mission that firstly orbited Earth and gathering data from 2002 to 2017. GRACE-FO will imitate the work of its antecedent while trying new technology adopted to vividly improve the already significant accuracy of its measurement system (NASA 2018).

Studies of water storage changes in hydrological reservoirs are expressed as equivalent water depth expressed in centimetres. There are three conventional methods for capturing and processing gravity data which include: the NASA Jet Propulsion Research Laboratory (JPL) method, the University of Texas Centre for Space Research (CSR) and the method of the GFZ centre. The various solutions of JPL, CSR, and GFZ can be applied to infer the uncertainty in the GRACE and GRACE-FO fields of Levels 2 and 3 that arises from the choice of the adopted strategy. Recent papers (Sakumura et al. 2014) found that mean average (simple arithmetic sum of JPL, CSR, GFZ fields) was effective in reducing gravity field noise in the available dispersion of all methods. It is therefore recommended to average the solutions of the data from the three centres (JPL, CSR, GFZ). For the case of the Mellegue watershed, only available valid data for the period 2002 until 2021 belong to the JPL and CSR centres. Therefore, an arithmetic average will be assigned based on data from both applied methods.

Global Land Assimilation System (GLDAS)

The purpose of GLDAS is to incorporate satellite and products of ground observational data, using innovative land surface modelling techniques and data assimilation, to generate optimal land surface states and fluxes. The software drives several offline land surface models (not coupled to the atmosphere), integrates a huge amount of observation-based data, runs globally at high resolution (2.5 degrees at 1 km), and is capable of producing results in near real time.

GLDAS-2 of NASA is the second version and it consists of three modules: GLDAS-2.0, GLDAS-2.1 and GLDAS-2.2. The first component (GLDAS-2.0) is fully coupled with weather data from Princeton and provides a time-consistent series that extends from 1948 to 2014. The second component GLDAS-2.1 constitutes a combination of model and observational data that began from 2000 to present. GLDAS-2.2 is the final product suites use data assimilation unlike the GLDAS-2.0 and GLDAS-2.1 products where they do not apply data assimilation. The choice of input data, observation source, variable and data assimilation scheme vary for various GLDAS-2.2 products. GLDAS-2.2 is recent to GES DISC (The NASA Goddard Earth Sciences (GES) Data and Information Services Centre (DISC)) archive and currently includes a main product of CLSM-F2.5 (Catchment Land Surface Model F2. 5) with data assimilation of the Gravity and Climate Recovery Experiment (GRACE-DA) data. GLDAS-2.2 data is available from February 2003 to present and offered two workflows: one with data assimilation outputs from GRACE (GRACE – DA) (the main workflow) and one without data assimilation from GRACE (the initial workflow). Since GRACE data has a latency of 2–6 months, GLDAS-2.2 data are first created by GRACE without assimilation and will be referred to as the primary product, with an approximate of one month latency. Once GRACE data are available, the GLDAS-2.2 data is processed with data assimilation from GRACE into the second workflow (main workflow) and will be removed from the primary product archive. The product GRACE-DA of GLDAS-2.2 was simulated with CLSM-F2.5 in the version 7 of Land Information System (LIS). The data product covers 24 fields of land surface and are available from 2003 to present. The simulation started on February 1, 2003 using the environmental settings of the GLDAS-2.0 Daily Catchment model simulation, coupled with the meteorological analysis fields of the operational integrated forecasting system of the ECMWF (European Centre for Medium-Range Weather Forecasts). The total observation of terrestrial water anomalies from the GRACE satellite was assimilated. Due to the data agreement with the ECMWF, this daily GLDAS-2.2 product does not include meteorological data fields. GLDAS-2.2 data is stored and distributed in NetCDF format.

Shallow groundwater storage simulated by the CLSM in GLDAS is defined as the water stored below the soil layer of the root zone and above the bedrock. Shallow groundwater storage is isolated from other water storage variables simulated by CLSM using this formula:

$$GWS=TWS-RMS-SWE-CI$$

With:

GWS: shallow groundwater storage; TWS: storage in terrestrial waters; RSM: root zone soil moisture, SWE: snow water equivalent; CI: canopy interception.

The version of the Noah land surface model currently used in NASA's GLDAS does not have a groundwater component. However, Noah-MP, which is used in other LDAS projects including Western LDAS, simulates GWS in a separate layer below the root zone.

Population census of Mellegue catchment

The following figures (Figs. 2, 3 and 4) shows the evolution of the demographic population of the Mellegue catchment area which was calculated on the basis of Worldpop data program. Referring to these figures, we notice that there is an urban growth at the level of large cities such as Tebessa (Algeria) and Kef (Tunisia). The most condensed urban agglomeration is concentrated in Tebessa with more than 10 000 inhabitants/km (Fig. 4). We obviously observe a demographic increase on the Tunisian-Algerian border marked by the birth of new urban agglomerations near the region of Tajerouine (the governorate of Kef) with a demographic density lower than 1000 inhabitants/km. In addition, we observe a reduction in demographic occupancy rates in almost all urban areas that are part of Tunisian territory. The only exception lies in the city of Kef where we note that there is a demographic increase in the centre of the city resulting in 4,000 and up to 10000 inhabitants/km locally in 2020 against 2,000 inhabitants/km in 2002, which requires development and urban extension to retain this population growth. In addition, the Tunisian part was obviously marked by the installation of new urban areas manifested by low populations limited to 1000 inhabitants/km. The installation of these new agglomerations automatically implies a probable urban extension.

On the other hand, the Algerian part of the watershed shows a considerable increase in population density observed in several regions such as the city of Ouenza, Meskiana and Khenchela. The regions of Tebessa and M'Daourouch are the most affected by population growth where the demographic occupancy rate exceeds 10,000 inhabitants/km (Fig. 4). This shows that the Algerian part of the basin has experienced a more advanced demographic development compared to the Tunisian territory which seems to have a low growth rate.

A previous research study (Weslati et al. 2022) proofs an extension of more than 3000 hectares of urbanized areas on the study area during 2002–2021. The general census of the population of the Mellegue watershed (Figs. 2 and 3, Table 1) shows an estimated population increase of 241837 in 2020 compared to 2002 with an average annual growth rate of 1.19%. The population census by country within the basin confirms the previous conclusions where we observe demographic stability in Tunisia throughout this period. Again, the region shows in some years a considerable reduction in population (2018/-1875). On the other hand, the same graphs show a significant increase of the Algerian population where it annually exceeds 10 000 inhabitants/year. It should be noted that the Tunisian regions of the watershed (Governorate of Jendouba, Kef and Kasserine) are considered among the poorest governorates of the country. School drop-out and unemployment rates are the highest (Dhraief et al. 2020). The economic crisis and the absence of stable employment opportunities promotes the development of informal trades. All these factors have imposed a migration from these regions to the eastern coast because of the economic and social potentialities. The rate of migration was intensified considerably from 2011 (revolution) following socio-political crises of the state.

Table 1 Population Census in Mellegue Watershed from 2002 to 2020

Year	Total population in the Mellegue catchment	Rate of population growth (inhabitant)	Rate of population growth (%)	Tunisian population in the bassin	Rate of population growth (inhabitant)	Rate of population growth (%)	Algerian population in the bassin	Rate of population growth (inhabitant)	Rate of population growth (%)
2002	1 124 852	0	0,00%	331 869	0	0,00%	792 984	0	0,00%
2003	1 130 945	6093	0,54%	328 231	-3638	-1,10%	802 714	9731	1,23%
2004	1 142 782	11 837	1,05%	326 659	-1572	-0,48%	816 123	13 409	1,67%
2005	1 147 153	4370	0,38%	322 851	-3808	-1,17%	824 301	8178	1,00%
2006	1 159 648	12 495	1,09%	325 131	2279	0,71%	834 517	10 216	1,24%
2007	1 167 464	7816	0,67%	322 936	-2194	-0,67%	844 528	10 010	1,20%
2008	1 180 646	13 182	1,13%	321 788	-1148	-0,36%	858 858	14 330	1,70%
2009	1 209 711	29 064	2,46%	326 672	4883	1,52%	883 039	24 181	2,82%
2010	1 215 051	5340	0,44%	323 838	-2834	-0,87%	891 213	8174	0,93%
2011	1 235 205	20 154	1,66%	321 430	-2408	-0,74%	913 774	22 562	2,53%
2012	1 246 557	11 352	0,92%	321 472	41	0,01%	925 085	11 310	1,24%
2013	1 263 801	17 244	1,38%	322 442	970	0,30%	941 358	16 274	1,76%
2014	1 280 222	16 422	1,30%	321 964	-478	-0,15%	958 259	16 900	1,80%
2015	1 293 120	12 898	1,01%	321 036	-928	-0,29%	972 084	13 825	1,44%
2016	1 312 370	19 250	1,49%	323 685	2648	0,82%	988 685	16 601	1,71%
2017	1 326 584	14 214	1,08%	323 323	-362	-0,11%	1 003 261	14 576	1,47%
2018	1 339 768	13 184	0,99%	321 448	-1875	-0,58%	1 018 320	15 059	1,50%
2019	1 355 466	15 698	1,17%	324 229	2781	0,87%	1 031 237	12 917	1,27%
2020	1 366 689	11 223	0,83%	324 373	144	0,04%	1 042 316	11 079	1,07%
Total	1 366 689	241 837	21,50%	324 373	-7495	-2,26%	1 042 316	249 332	31,44%

Forest investigation

Forests are the concern of the whole world; they are threatened by irresponsible deforestation and fires. Figure 5 shows NDVI interpreted map based on Landsat images for 2002, 2012, 2018 and 2022. It clearly shows a net degradation of the forests especially in the downstream part of the watershed. Other additional studies (Weslati et al. 2022) have reported the loss of more 15 000 hectares of forests in the catchment of Mellegue during the period 2002–2021. This is explained by several effects; in Tunisia, the socio-political status of the country since 2011 (revolution) has seriously affected the forestry sector. Consequently, the deforestation index has increased, but it was not calculated in the latest FAO statistics due to these socio-economic problems. Still, more than 400,000 m³ of wood has been produced from Tunisian forests (Chriha and Sghari 2013). In return, deforestation in Algerian regions has also intensified, registering an increase from 11,000 ha/year in 2005 to more than 18,000 ha/year in 2018 (FAO 2015a; FAO 2015b).

Always coveted by the population, forest areas constitute a means of subsistence where various local daily activities are organized around them such as the production of charcoal, the collection of firewood or even artisanal wood and cork activities. Unfortunately, forests are still over-exploited causing considerable environmental and ecological degradation. In Tunisia, the exploitation of forest products is controlled and regulated through a specialized dedicated department of the Ministry of Agriculture. Figure 6 and Table 2 summarize a statistic of the number of forest offenses legally proceeded for twenty years (between 2001 and 2021) and which took place in the Tunisian part of the Mellegue watershed. They reveal that the number of fines in 2018 was multiplied five times compared to 2002 (51 in 2002 against 251 in 2018). It is also noted that the number of offenses during the first decade (from 2001 to 2010) is almost stable. The turning point was 2011 following the socio-political unrest that lasted throughout the year. At that time, the state lost much of its authority where only 25 offenses were written and which required a serious intervention of the army to struggle against these acts of ecological degradation. Since then, even if the state has restored its control, the number of reported offenses has continued to increase, exceeding 350 fines in 2021 compared to only 63 in 2001.

Secondly, fires remain by far the primary threat due to their ability to quickly ravage large areas. According to the results of the historical fires from the MODIS and VIIRS sensors (Fig. 7 and Table 2), the impact of the fires seriously affected the vegetation in the Mellegue catchment. The number of fires increased tragically during the second decade (2011–2021) compared to the first (2002–2011). For example, the number of annual fires in 2013 is almost equal to the sum of all the fires recorded during the first decade (2002–2011) with a total of 144 fires.

Table 2

Statistics of *fires number occurring in the catchment of Mellegue forests during the period 2002–2021*
Year	Algeria	Tunisia	Total
2002	1	0	1
2003	6	3	9
2004	10	2	12
2005	14	3	17
2006	6	3	9
2007	6	6	12
2008	31	1	32
2009	23	3	26
2010	10	1	11
2011	13	2	15
2012	33	57	90
2013	27	29	56
2014	50	165	215
2015	13	65	78
2016	47	86	133
2017	29	109	138
2018	19	18	37
2019	49	63	112
2020	24	69	93
2021	64	57	121

However, the thorough evaluation of this figure allowed to classify the number of fires by country. The socio-political instability of Tunisia since 2011 has impacted the vegetative spaces. From 2012, the number of fires begins to increase intensely compared to the antecedent years. The record took place in 2014 with 215 fires in the entire basin where 80% of them (169 fires) are started in Tunisia. During this year, a record was reached in terms of burned areas, indicating more than 6000 ha of land that were invaded by these fires (Fig. 8). According to these results, it can be seen that the period between 2012 and 2014 led to an unprecedented resurgence of forest fires where Tunisia was punctuated by multiple political, terrorist and social crises. Although the state has regained its power where a certain stability has been felt, leading to a drop in the number of fires, especially in 2018 (37 in total with 18 in Tunisia), the values remain high compared to the first period (2002–2011). Over the entire basin, the sum of the fires accumulated in the second decade (2012–2021) is equal to 1073 where 718 of these fires are initiated in Tunisia against 355 in Algeria. The number is multiplied 10 times compared to the period 2002–2011 with only 144 fires (24 fires in Tunisia and 120 in Algeria) reported in the entire basin.

The Algerian territory recorded rhythmic fires with a slight increase compared to the first decade. The political instability of Tunisia may have affected the neighbouring forest areas of Algeria. Some fires were deliberately started in Tunisia and invaded the Algerian forests (or vice versa.). Forest areas are always exposed to climate change, especially during years of drought, but significantly to anthropogenic actions. In conclusion, the forests of the Mellegue watershed have been the victim of large-scale degradation. For example, since 2011, more than 1 000 ha of the forests of the Djebel Ouergha mountain under the delegation of Sakiet Sidi Youssef have been devastated by fires following anthropogenic, climatic and terrorist actions (Fig. 9).

Deep water analysis and quantification

In-depth analysis prove that water resources of the basin are seriously threatened. To begin with, the area witnessed an expansion of agricultural land and has thus led to an increase in water requirements for irrigation. In Algeria, more than 500 000 ha were added to agricultural land in 2014 compared to 1994 (Harbouze et al. 2019). Consequently, the consumption of surface or deep water has reached a very high threshold. To compensate this excessive demand for water, the number of dams has increased from 44 with a storage capacity of 3.3 billion m³ in 1994 to 100 dams in 2016 with a mobile water storage of 12 billion m³. Half of this water is dedicated to agricultural irrigation (6 billion m³). As a result, the area of irrigated land increased from 350 000 ha in 2000 to 1.3 million ha in 2017 (Harbouze et al. 2019). On the other hand, it should be noted that Tunisia suffers from a shortage of water due to the reduction in lakes and surface reservoirs following the intensification of the arid climate. The area of irrigated land has increased from 380 000 ha in 2001 to 460 000 ha in 2010, which has resulted to an increase in the need for water resources for irrigation (Chebbi et al. 2019). In addition, Tunisian regions of the basin are mainly based on agriculture and associated activities. For example, 25% of the population of the governorate of Kef is absorbed by the agronomic sector where the agricultural area occupies almost 70% of the region total area. Cereal farming is the main activity with 43% of the occupation rate compared to the total surface of the governorate. The installation of several irrigated areas and wells has enabled the development of the cereal sector and plant nurseries. In 2017, the governorate of Kef disposed already 43 000 hectares of irrigated areas compared to 10 000 hectares in 2005 (Samaali and Mjejra 2020). In addition, the Tunisian ministry of agriculture announces the installation of a new irrigated areas with an estimated surface area of 2000 hectares which will be supplied with water from the new Mellegue 2 dam. As for Jendouba (outlet of the catchment), it participates in the national production of various products, particularly in cereal production and market gardening where it also has more than 40 000 hectares of irrigated areas (Chebbi et al. 2019). The Algerian territory of the catchment also invests in the agricultural sector. The region of Tebessa is largely based on agriculture, it has nearly 20 000 hectares of irrigated perimeters. The governorate announced in 2014 a deduction of a new irrigated area exceeding 30 000 hectares. The new area will be fed by the Ouldjet Mellegue dam and two small dams planned at Ferkane and Ouenza (Azzeddine 2011). These facts have led to an excessive use of shallow and deep waters, which has modified the natural characteristics of water cycles and resulted in a significant degradation of the quality of these waters mainly by salinization (Chebbi et al. 2019; Laajimi 2011).

The geographical location, as well as the climatic characteristics, classify Tunisia and Algeria under high-water stressed areas (500 m³/inhabitant/year) (Hofste et al. 2019). It can also be expected that the demand for the supply of water resources will be intensified in proportion to the current population growth. Mellegue dam in situated in the downhill of the catchment, in El-kef near the outlet. The location of the dam made it possible to supply the approximate towns with drinking water and be used for irrigation. However, in recent years, the dam has suffered a sharp decrease in its waters. Figure 10 shows an NDWI-based comparison of the dam’s water between 2002, 2012 and 2022. A serious quantitative drop in the water resources of the dam can be clearly be notified. Figure 11, followed by Table 3 illustrates the difference in the contour of the water retention of the dam in 2002 compared to 2022 which clearly proves the decrease in the water reserves of the dam. Indeed, the area of the dam changed from 600 ha in 2002 to 320 ha in 2012 until reaching 242 ha in 2022, while the perimeter of water retention decreased from 55.5 km in 2002 to 19.8 km and 16.6 km for 2012 and 2022 respectively.

The natural muddy flows of the main river have caused the siltation of more than half of the dam storage. Some study claimed that the dam would be completely silted up by 2025 (Cherni et al. 2010). The water become too salty reaching 10 g/l. Another ancillary study reported that the dam's water storage is actually around 5 million m³ where they recorded more than 68 million m³ in 2002 (Weslati et al. 2020). In fact, this decrease was partially due to squandering of the dam's water resources for the irrigation and drinking water, dropping from 292 million m3 in 2002 to less than 6 million m³ actually.

Table 3

*Water storage contrast in the Mellegue dam for 2002, 2012 and 2022*
Date of observation	December-2002	December-2012	December-2022
Instrument	Landsat ETM+	Landsat ETM+	Landsat OLI
Surface of water dam (hectare)	600	320	242
Perimeter of water dam (kilometre)	55.5	19.8	16.6

The approximate regions of the Mellegue dam are increasingly suffering from a shortage of water manifested by a disruption and a daily cut in the distribution of drinking water. Considerable efforts have been made to manage the surface waters of the basin; in 2016, the Tunisian state started construction work on the upper Mellegue dam. The reservoir of the dam has a surface of 1320 ha with a maximum reservoir capacity estimated at 195 million m³. The average annual supply is estimated at 170 million m³, of which 87 million m³ of water will be transported to the Sidi Salem dam while the rest (76 million m³) will be mobilized for the existing perimeters (drinking water and irrigation) with 8.7 million m³/year dedicated to the creation of a new irrigated perimeter estimated with an area of 2000 ha. Works were also been carried out to fight against floods which frequently occur near the outlet of the Mellegue catchment (in the plain of Jendouba) as well as to produce electricity (Fehri 2014; Sahar et al. 2019; Université de Tokyo 2013). In return, the Algerian state completed the execution of the Ouldjet Mellegue dam in 2019. It is located upstream of the basin, 15 km from the city of Ouenza and 75 km from Tebessa. The maximum capacity of the reservoir is estimated at 156 million m³. The development of the dam is intended for the mobilization of water from the main river where approximately 130,000 inhabitants of neighbouring municipalities will be supplied with reservoir water. Part of the dam’s water is also planned to meet the needs of the phosphate complex installed in the town of Oued Kerbrit. Similarly, the objective of the development of the dam is envisioned to meet the irrigation water needs of agricultural land in the region of Ouenza.

The geographical location of Tunisia classifies it as a country with an arid to semi-arid climate over almost all of its surface with low to medium rainfall. Indeed, Tunisia has a low volume of available water estimated at 500 m³/inhabitant. A rate that ranks the country among the regions with high water stress (Solanke et al. 2005). Again, these water resources are poorly distributed where northern of Tunisia is responsible for generating 80% of water surface and 60% of the overall water contribution to the country (Cherif 2003). The arid to semi-arid climate has aggravated the situation due to the inter-annual variability of rainfall which generally brings low precipitation rates.

Secondly, the average water consumption for Tunisia is estimated at 70 litres per resident per day in urban areas compared to 50 litres per resident per day in rural areas (WMC and TAP 2019). The forecast of future water consumption for Tunisia tends to double around the 2050 horizon, estimated at 140 litres per inhabitant per day in urban areas against 100 litres per resident per day in rural areas (WMC and TAP 2019). The expansion of urban areas in the watershed of the Mellegue will necessarily cause an increase in water consumption, amplifying the pressure on water resources which requires recovery and management actions integrating not only this basin but all the watersheds of Tunisia.

Despite a narrow surface area and rigid climatic factors, Tunisia claims a considerable potential in thermo-mineral water. The bottled water market in Tunisia is growing steadily. In one decade, consumption has increased from 290 million litres in 2000 to 900 million litres in 2009, reaching 2.7 billion litres in 2020. These statistics rank Tunisians among the biggest consumers of mineral water in the world (4th rank according to (research and markets 2022) To meet daily needs, there has been an increase in the number of production units. Currently, 29 units are installed in 12 governorates producing more than 364 000 bottles per hour. The number of units was expanded 4 times compared to 1994 with only 7 packing plants (Oueslati 2021).

In this context, the packaging of bottles of mineral water is done in and near the watershed of the Mellegue thanks to an aquifer located on its NW edge. In this part, there are three mineral water conditioning plants (figure). Two factories belonging to the same company group (SOSTEM) dating from 1968 (Safia Aïn Mizeb) and 1989 (Safia Aïn Ksiba) for the packaging of mineral water from a limestone aquifer of Ypresian age (brand “Safia”). The group's statistics confirm an increase in production resulting in 130 million bottles in 2021 compared to 68 million bottles in 2006 from the Aïn Mizeb source (Fig. 13). The second unit (Aïn Ksiba) confirms an annual production of 223 million bottles in 2021 compared to 147 million bottles in 2013 (Fig. 13). This will most likely lead to a depletion of deep-water resources in the region. The parent company has been carrying out exploration work on other wells for years. Currently, it has several boreholes to supply their production units. The depletion of the aquifer's water resources intensifies when a third unit based in the city of Tajerouine was installed to exploit the waters of a sand-gravel aquifer of Mio-Pliocene age, dedicated for conditioning brand bottled mineral water (brand “Dima”). The factory was installed in 2009 with a production capacity rate of 40 000 bottles per hour.

The results of the variation of the surface and subsurface waters of the basin were interpreted by GRACE and GRACE-FO satellite data. According to these results illustrated in Fig. 14, we notice the existence of two methods from the JPL (red) and CSR (orange) labs to which is associated a curve representing the arithmetic mean of the two methods (blue). However, there is a slight difference between the two applied methods (JPL and CSR). The results of the JPL lab show a weak to stable variation in the period 2002–2010. From 2011, the decline increases continuously until reaching in 2021 a reduction of -20 cm in equivalent water depth.

CSR's research centre results follow almost the same pattern of JPL during the first period (2002–2010) with a slight variation. The difference stands out in the second period (2011–2021) where the decrease is less pronounced. The general trend of the curve indicates a reduction in resources intercepted by a few recharge periods. The general balance of the CSR curve shows a reduction of about − 10 cm in water equivalent thickness during the period of 2002–2021.

From the average of these two results (blue), it can be seen that the behaviour of the equivalent water thickness was influenced by two different periods. The first period (2002–2010) was marked by a certain stability manifested by repetitive recharge periods where peaks exceeding + 5 cm in equivalent water thickness are often observed. From 2011, these peaks are no longer observed throughout the following period. There is a continuing and deeper decline. The recharge periods are not very spread out and of very short duration. With the exception of the event that took place in March 2012, all recharge periods observed barely crossed the bar of + 0 cm of equivalent water thickness.

It can be seen that the resources of the basin are overexploited in the second decade (2011–2021) where the trend curve (black) shows a decreasing with continuous pace, which can be linked to the socio-political events in Tunisia dating from 2011 so far. If we take the results of the trend curve interpreted annually on the average of the results of the two research centres (JPL and CSR), we deduce a reduction of about − 16 cm in equivalent thickness water where it has been intensely declined during the period 2011–2021. Total estimates claim an overall exploitation of approximately 1.7 billion m³ of the water resources of the basin (surface and deep).

Based on all these points, it is clear that the state of water resources in the Mellegue watershed is alarming mainly due to the excessive and anarchic consumption of water resources for the different sectors (irrigation, industry, sanitation, drinking water) and which intensified especially in Tunisia from 2011.

In order to monitor the groundwater reserves of the Mellegue watershed, GLDAS data provided by NASA made it possible to quantify the volume of groundwater throughout the entire basin. The data available is from 2003 to the present. The general situation shows a decrease in the available volume (Figs. 15 and 16) starting from an initial volume estimated at 11.7 billion m³ in 2003 to 9.5 billion m³ by the end of 2021. On the basis of these estimates, we deduce an approximate consumption of two billion m³ of water resources throughout the period, which coincides with the previous results confirmed by the GRACE data. Consequently, the annual balance sheet shows an average consumption of 120 million m³ per year. This work also shows that the Tunisian water reserves constitute about 50% of the total volume found in the basin. This implies that there has been an overexploitation of these waters, linked to strong demographic pressure and agricultural activities, which has caused a drop in the amount of available basin water.

The uncertainties associated with the input variables have repercussions, in terms of indeterminacy, on the results, whose conclusions stated above remain significant. However, it is necessary to implement deeper studies within the basin for a detailed assessment of water resources (deep and surface) for sustainable management. This requires a study of the mechanisms responsible for the quantitative and qualitative degradation of water. Knowing that good management requires a detailed understanding of the basin, requested in particular for similar regions characterized by an arid climate where the rainfall deficit weighs on the recharge of the aquifers.

Based on geospatial data and remote sensing analysis, the present study has exhibited a major degradation of Mellegue catchment resources. It highlighted a population growth which has led to certain extension of new urban areas. This was confirmed by demographic occupancy maps illustrating an increase in the population occupancy rate in the centres of several cities such as El-Kef region in Tunisia or Tebessa, Ouenza, M'Daourouch and Meskiana in Algeria. The maps show also a settlement of new communes with a population less than or equal to 1000 inhabitants/km in several provinces of the basin. The second main event consist in a significant reduction of forest area was caused by the combination of climatic and anthropogenic factors which were amplified due to the social, terrorist and political troubles following the 2011 revolution. In-depth quantification of water resources shows a considerable depletion of water resources (both surface and deep) that has been caused by overexploitation to meet agricultural, social and industrial needs where the consumption rate has been highly intense in the present decade following the 2011 revolution events. These threats require an increase in the intervention methods, in both preventive and curative. The protection of the catchment resources can only be effective through the establishment of appropriate and efficient ways of surveillance. The link with the precarious economic situation of the Tunisian population in the watershed as a result of “Jasmine revolution” is obvious. Without the rapid improvement in the living conditions of local residents, any attempt at protection is already doomed to failure.

This work calls for the integration of recent geospatial data to produce reliable information on the threats of the basin resources. This work will definitely open new perspectives based on the use of GIS tools and remote sensing for the management of resources at the scale of a hydrological basin. However, it is necessary to adapt these tools in front of the constraints and issues related to the management of such a large hydrographic basin like Mellegue catchment where the challenges can be locally high and the depletion of resources increases under the pressure of anthropogenic activities and the arid climate that dominates the region.

Author’s Contribution

All authors contributed to the study topic and research.
Material preparation, data collection and analysis were performed by Okba Weslati and Mohamed Moncef Serbaji.
The first draft of the manuscript was written by Okba Weslati.
The first written draft of the manuscript was reviewed and corrected by Mohamed Moncef Serbaji.
All authors read and approved the final manuscript.

Competing interests

The authors have no relevant financial or non-financial interests to disclose.

Availability of Data and Materials

Please address to Okba Weslati ([email protected]) for data request.

Funding

No funding was obtained for this study.

Azzeddine K. Extension attendue des superficies agricoles irriguées Tébessa. Le Maghreb. 2011;
Belloula M. Evaluation de l’aptitude aux écoulements et risque d’érosion dans le haut cours de la Medjerda par Modélisation. Institut des sciences de la terre et de l’univers; 2017.
Belloula M, Dridi H. Modeling of the flows and solid transport in the catchment area of Meskiana-Mellegue upstream (Northeastern Algeria). Geogr. Tech. 2015;10(1):1–7.
Bhandari AK, Kumar A, Singh GK. Feature Extraction using Normalized Difference Vegetation Index (NDVI): A Case Study of Jabalpur City. Procedia Technol. 2012;6:612–21.
Bucała A. The impact of human activities on land use and land cover changes and environmental processes in the Gorce Mountains (Western Polish Carpathians) in the past 50 years. J. Environ. Manage. 2014;138:4–14.
de Carvalho DF, Durigon VL, Antunes MAH, de Almeida WS, de Oliveira PTS. Predicting soil erosion using Rusle and NDVI time series from TM Landsat 5. Pesqui. Agropecu. Bras. 2014;49(3):215–24.
Chebbi HE, Pellissier J-P, Jean-Pierre Rolland, Wided Khechimi. Rapport de synthèse sur l’agriculture en Tunisie [Internet]. Ciheam-Iamm. 2019. Available from: https://hal.archives-ouvertes.fr/hal-02137636
Cherif A. Le problème de l’eau en Tunisie nord-orientale: besoins, ressources locales et transferts inter-régionaux. EAU Environ. Tunisie milieux méditerranéens. 2003. p. 208.
Cherni S, Khlifi S, Louati MH. SUIVI DE L’ENVASEMENT DE LA RETENUE DU BARRAGE DE NEBEUR SUR L’OUED MELLEGUE (LE KEF). Actes des 17èmes Journées Sci. sur les Résultats la Rech. Agric. 2010.
Chriha S, Sghari A. Forest fires in Tunisia, irreversible sequelae of the revolution of 2011. J. Mediterr. Geogr. 2013;121.
Deshayes M, Guyon D, Jeanjean H, Stach N, Jolly A, Hagolle O. The contribution of remote sensing to the assessment of drought effects in forest ecosystems. 2020;
Dhraief D, Boulila G, Pavelesku D, Gannat HM. Carte de la pauvreté en Tunisie. 2020.
EOS Data Analytics I. NDVI: Normalized Difference Vegetation Index. EOS Data Anal. 2023. p. 1.
FAO. EVALUATION DES RESSOURCES FORESTIERES MONDIALES 2015 - Algérie. 2015a;.
FAO. EVALUATION DES RESSOURCES FORESTIERES MONDIALES 2015 - TUNISIE. 2015b;.
Fehri N. L’aggravation du risque d’inondation en Tunisie: éléments de réflexion. Physio-Géo. 2014;(Volume 8):149–75.
Fraihat I. Unfinished Revolutions: Yemen, Libya, and Tunisia after the Arab Spring. Yale University Yale; 2016.
Gao BC. NDWI - A normalized difference water index for remote sensing of vegetation liquid water from space. Remote Sens. Environ. 1996;58(3):257–66.
Gaughan AE, Stevens FR, Huang Z, Nieves JJ, Sorichetta A, Lai S, et al. Spatiotemporal patterns of population in mainland China, 1990 to 2010. Sci. Data. 2016;3:1–11.
Giglio L, Descloitres J, Justice CO, Kaufman YJ. An enhanced contextual fire detection algorithm for MODIS. Remote Sens. Environ. 2003;87(2–3):273–82.
Gu Y, Brown JF, Verdin JP, Wardlow B. A five-year analysis of MODIS NDVI and NDWI for grassland drought assessment over the central Great Plains of the United States. Geophys. Res. Lett. 2007;34(6):1–6.
Gu Y, Hunt E, Wardlow B, Basara JB, Brown JF, Verdin JP. Evaluation of MODIS NDVI and NDWI for vegetation drought monitoring using Oklahoma Mesonet soil moisture data. Geophys. Res. Lett. 2008;35(22):1–5.
Harbouze R, Pellissier J-P, Rolland J-P, Khechimi W. Rapport de synthèse sur l’agriculture en Algérie [Internet]. 2019. Available from: https://hal.archives-ouvertes.fr/hal-02137637/document
Hill MJ, Donald GE. Estimating spatio-temporal patterns of agricultural productivity in fragmented landscapes using AVHRR NDVI time series. Remote Sens. Environ. 2003;84(3):367–84.
Hofste RW, Reig P, Schleifer L. 17 Countries, Home to One-Quarter of the World’s Population, Face Extremely High Water Stress. World Resour. Inst. 2019;1.
Illera P, Fernández A, Delgado JA. Temporal evolution of the NDVI as an indicator of forest fire danger. Int. J. Remote Sens. 1996;17(6):1093–105.
J.A.Rodier, J.Colombani, J.Claude, R.Kallel. Le Bassin de la Mejerdah [Internet]. 1981. Available from: https://www.worldcat.org/title/bassin-de-la-mejerdah/oclc/469086740
Krtalic A, Prodan A, Racetin I. Analysis of burned vegetation recovery by means of vegetation indices. Int. Multidiscip. Sci. GeoConference Surv. Geol. Min. Ecol. Manag. SGEM. 2019. p. 449–56.
Laajimi A. Les périmètres irrigués en Tunisie Un enjeu pour le développement de la production agricole. 2011.
Landerer FW, Cooley SS. GRACE D-103133 Gravity Recovery and Climate Experiment Level-3 Data Product User Handbook [Internet]. 2019. Available from: https://podaac-tools.jpl.nasa.gov/drive/files/allData/gracefo/docs/GRACE-FO_L3_Handbook_JPL-D-103133_20190327.pdf
Liping C, Yujun S, Saeed S. Monitoring and predicting land use and land cover changes using remote sensing and GIS techniques—A case study of a hilly area, Jiangle, China. PLoS One. 2018;13(7):1–23.
Mahmoud SH, Gan TY. Impact of anthropogenic climate change and human activities on environment and ecosystem services in arid regions. Sci. Total Environ. [Internet]. Elsevier B.V.; 2018;633:1329–44. Available from: https://doi.org/10.1016/j.scitotenv.2018.03.290
Martine G, Schensul D. The Demography of Adaptation to Climate Change [Internet]. Demogr. Adapt. to Clim. Chang. 2013. Available from: http://www.unfpa.org/sites/default/files/pub-pdf/The Demography of Adaptation to Climate Change.pdf
Mlayah A, Ferreira da Silva E, Rocha F, Hamza C Ben, Charef A, Noronha F. The Oued Mellègue: Mining activity, stream sediments and dispersion of base metals in natural environments, North-western Tunisia. J. Geochemical Explor. [Internet]. Elsevier B.V.; 2009;102(1):27–36. Available from: http://dx.doi.org/10.1016/j.gexplo.2008.11.016
Mlayah A, Ferreira Da Silva EA, Lachaal F, Khadhar S, Charef A, Noronha F. Effet auto-épurateur de la lithologie des affleurements géologiques dans un climat semi-aride: cas du bassin versant de l’Oued Mellègue (Nord-Ouest de la Tunisie). Hydrol. Sci. J. 2013;58(3):686–705.
Mlayah A, Lachaal F, Chekirbane A, Khadar S, da Silva EF. The Fate of Base Metals in the Environment and Water Quality in the Mellegue Watershed, Northwest Tunisia. Mine Water Environ. [Internet]. Springer Berlin Heidelberg; 2017;36(2):163–79. Available from: http://link.springer.com/10.1007/s10230-017-0430-z
NASA. Recovery, Gravity Follow-on, Climate Experiment. 2018;
NASA. Fire Information for Resource Management System (FIRMS) [Internet]. 2022. Available from: https://firms.modaps.eosdis.nasa.gov/
Oueslati R. Le Guide du Consommateur des Eaux Embouteillées en Tunisie. Office national du thermalisme et de l’hydrothérapie; 2021.
Qiao C, Luo J, Sheng Y, Shen Z, Zhu Z, Ming D. An Adaptive Water Extraction Method from Remote Sensing Image Based on NDWI. J. Indian Soc. Remote Sens. 2012;40(3):421–33.
Rawat JS, Kumar M. Monitoring land use/cover change using remote sensing and GIS techniques: A case study of Hawalbagh block, district Almora, Uttarakhand, India. Egypt. J. Remote Sens. Sp. Sci. [Internet]. Authority for Remote Sensing and Space Sciences; 2015;18(1):77–84. Available from: http://dx.doi.org/10.1016/j.ejrs.2015.02.002
research and markets. Bottled Water Manufacturers (Global) - Industry Report. 2022.
Rouvier H. Géologie de l’extrême nord-tunisien: tectonique et paléogéographie superposées à l’extrémité orientale de la chaine nord-maghrébine. Paris VI; 1977.
Sahar A, Olfa H, Ezzine A, Hermassi T. Cartographie du risque d ’ inondation de la Haute Vallée de Medjerda avec Hec- Ras. 2019;(January 2020).
Sakumura C, Bettadpur S, Bruinsma S. Ensemble prediction and intercomparison analysis of GRACE time-variable gravity field models. Geophys. Res. Lett. 2014;41(5):1389–97.
Samaali H, Mjejra M. Essai typologique des périmètres irrigués du Gouvernorat d ’ El Kef (Nord-ouest de la Tunisie): Un outil pour diagnostiquer et juger les performances des exploitations agricoles. Rev. Marocaine des Sci. Agron. Vétérinaires. 2020.
Singh D, Herlin I, Berroir JP, Silva EF, Meirelles MS. An approach to correlate NDVI with soil colour for erosion process using NOAA/AVHRR data. Adv. Sp. Res. 2004;33(3):328–32.
Small C, Sousa D. Humans on Earth: Global extents of anthropogenic land cover from remote sensing. Anthropocene [Internet]. Elsevier B.V.; 2016;14:1–33. Available from: http://dx.doi.org/10.1016/j.ancene.2016.04.003
Solanke PC, Srivastava R, Prasad J, Nagaraju MSS, Saxena RK, Barthwal AK. APPLICATION OF REMOTE SENSING AND GIS IN WATERSHED CHARACTERIZATION AND MANAGEMENT. J. Indian Soc. Remote Sens. 2005;(2).
Sorichetta A, Hornby GM, Stevens FR, Gaughan AE, Linard C, Tatem AJ. High-resolution gridded population datasets for Latin America and the Caribbean in 2010, 2015, and 2020. Sci. Data. 2015;2:1–12.
Stevens FR, Gaughan AE, Linard C, Tatem AJ. Disaggregating census data for population mapping using Random forests with remotely-sensed and ancillary data. PLoS One. 2015;10(2):1–22.
Université de Tokyo. Tunisie Etude Préliminaire du Projet Intégré de Gestion du Bassin Versant et de Contrôle des Inondations de La Mejerda: Analyse de l’Impact du Changement Climatique -RAPPORT FINAL. 2013.
Weslati O, Bouaziz S, Sarbeji MM. Modelling and Assessing the Spatiotemporal Changes to Future Land Use Change Scenarios Using Remote Sensing and CA-Markov Model in the Mellegue Catchment. J. Indian Soc. Remote Sens. [Internet]. Springer India; 2022;4. Available from: https://doi.org/10.1007/s12524-022-01618-4
Weslati O, Bouaziz S, Serbaji MM. Mapping and monitoring land use and land cover changes in Mellegue watershed using remote sensing and GIS. Arab. J. Geosci. Arabian Journal of Geosciences; 2020;13(14).
WMC, TAP. A l’horizon 2050, chaque Tunisien doublera sa consommation d’eau par rapport à 2014 (Etude). webmanagercenter [Internet]. 2019; Available from: https://www.webmanagercenter.com/2019/11/15/441344/a-lhorizon-2050-chaque-tunisien-doublera-sa-consommation-deau-par-rapport-a-2014-etude/
worldpop. Open Spatial Demographic Data and Research [Internet]. worldpop. 2013. Available from: https://www.worldpop.org/

No competing interests reported.

Application of Remote sensing and GIS on studying the impact of « Jasmin revolution » on Mellegue catchment

Status:

Version 1

Abstract

Figures

Introduction