## Flood viewer and inundation zone

As can be seen from the maximal flood levels in the delta over the years, the flooded surface area was at its largest in the wet periods in the 1920s and 1960s. The topographical maps of the Institut Géographique National (IGN) from that period reveal that the maximal inundation area of the Inner Niger Delta measures 36 470 km^{2}, including 5340 km^{2} of levees, dunes and other islands within that area. They also show that water coverage declines from 31 130 km^{2} in wet periods to 3840 km^{2} in the dry period. The entire floodplain area is included in the 41 195 km^{2} designated as a Ramsar Wetland Site of International Importance in January 2004. Topographical maps show the floodplain as if it were flooded at a maximum level (see **maximum flood viewer**). The area actually inundated, however, varies considerably between years, as can be seen in the previous Section.

The **maximum flood viewer** below shows the flood extent at a chosen water level, during upcoming water (crue). You can choose a water level between 0 and 520 (max. flood level in Akka), and the corresponding flood extent during that season is shown on the map.

Satellite images can be used to produce water maps of the Inner Niger Delta covering the full range of water levels. That is why we know that in 1984 (at a maximal flood level of 336 cm in Akka) a mere 7800 km^{2} was flooded. In contrast, 18 000 km^{2} was flooded in 1999, when the water level at Akka reached the high level of about 511 cm. Zwarts & Grigoras (Zwart et al. 2005) used satellite images to produce a digital flooding model of the Inner Niger Delta. The resulting spatial flooding patterns closely reflect the real inundation at a certain water level. The flooding model relies on the input of discharge or water levels. In 2016-2019 a new flooding model is build, covering the entire delta and based on 273 satellite images during upcoming water (Davids et al. 2019). The model separates water on the floodplain which are connected to the river and which are not connected (most depressions with rainfall, not (yet) flooded). With this model for each water level the spatial flooding pattern during incoming water can be established.

## Actual flooding

The annual flood extent is the driver of many socio-economic and ecological functions in the delta, such as the production of fish, rice and bourgou. Information on the expected timing and extent of flooding is of interest to many stakeholders. At present, monitoring of flooding can be done through recent satellite images, but these still need processing to translate them into actual water maps, but it might be a promising tool in the near future. Flooding can also be monitored indirectly through water levels. The flood in the Inner Niger Delta is forecasted each flooding season using the tool OPIDIN.

** OPIDIN** stands for ‘

*’ and is based on a model in which flooding characteristics are related to water levels in the delta. With the (automatic) input of rainfall data and actual water levels in Mopti, the tool forecasts the timing and maximal spatial flooding in the delta at the peak of the flood. The forecast is done annually during the flooding season, and is updated every week on the*

__O__util de__P__rédiction des__I__nondations dans le__D__elta__I__ntérieur du__N__iger**website. On this website also all background information on OPIDIN can be found as well as the actual predictions on flooding.**

__OPIDIN__The following information is given on the OPIDIN site:

**Maximum inundation level**Starting each year on August 1^{st}the maximum flood level is predicted on a weekly basis for three stations in the delta: Mopti, Akka and Diré. The expected level is given and the confidence range, expressed as a lower and upper limit. The forecast is stopped when the maximum flood level in Mopti has been reached in October.**Timing of flooding**

Next to the__height of the maximum flood__also the timing is predicted. For each of the three stations the expected date (and lower and upper range) is given at which the peak flood level is reached.**Spatial extent of flooding at time of flood peak**

The__spatial extent of inundation__during the expected flood peak is presented on an interactive map. This is based on the flooding model explained elsewhere in this Section.**Deflooding**

As soon as the peak flood is reached, the__timing of deflooding__is important to stakeholders. This is predicted for the level of 200 cm at the Akka gauge (below this level the first sand banks in the central lakes become exposed), and the level of 100 cm at the Akka gauge (most mud banks exposed).

## Annual inflow into the Inner Niger Delta

The Inner Niger Delta is flooded due to the large discharge of the Niger and the Bani during and after the rainy season. To estimate the annual inflow into the Inner Niger Delta, the daily measurements of the water level (measured in cm on the local gauge) have to be converted into daily discharge (m^{3}/s) using measurements at a wide range of water levels. The relationship between water level and river flow is not linear and DNH uses third-degree polynomials to get a good fit (Brunet-Moret *et al*. 1986; Zwarts & Diallo 2002).

The maximal flood level in the Inner Niger Delta is due to the discharge of the Niger and the Bani in the months before the peak is reached, thus still during incoming water which is for the Inner Niger Delta, on average, about 15 November. The annual variation of the inflow during incoming water is shown for three stations in the Bani.

*The total annual flow of the Bani during incoming water (1 June – 15 November) measured at three stations.*

Douna which was always the main hydrological stations for the Bani cannot be used anymore since the construction of the Talo dam downstream of Douna. Béneny-Kegni can be used very well instead, however, given the close relationship between the annual inflow for both stations:

*Bénény-Kegni (km ^{3}/year) = 0.9135 * Douna (R^{2}=0.989, n=52, calculated for the years 1952-2005). *

In contrast, Sofara deviates from Douna and Bénény-Kegni. As the graph shows the discharge is the same for both stations in dry years when the total discharge of the Bani is lower than 7 km^{3}.

*The annual discharge in Sofara as a function of the discharge in Bénény-Kegni in the same year, both calculated for the same period (1 June – 15 November). The straight line gives the expected relationship if both would be equal. If the total discharge of the Bani is larger than 7 km ^{3}, an increasing fraction of the water does not reach Sofara, because more and more water from the Bani River flows before Sofara to the Niger River, from Djenné to Kouakourou along the Manga tributary.*

The total annual inflow of Bani and Niger into the Inner Niger Delta equals the sum of the flow as measured at Macina for the Niger and at Bénény-Kegni for the Bani (see graph below).The flow of the Niger is, on average, four times larger than those of the Bani.

*The total annual flow at incoming water (1 June – 15 November) of the Niger (Macina) and the Bani (Bénény-Kegni).*

The Bani flows just before Mopti into the Niger. Thus, the gauge at Mopti measures the combined flow of Niger and Bani. Indeed, the trend at Mopti resembles those at Macina and Bénény-Kegni, with high flows in the 1950s, declining flows until 1990s and partial recover since then. However, the total flow at Mopti is lower than Macina and Bénény-Kegni and even lower than Macina alone.

*The total annual flow at incoming water (1 June – 15 November) of the Niger and Bani combined at Mopti.*

A separate graph shows the relationship between Mopti and Macina + Bénény-Kegni. In dry years, 68% of the flow of the Niger and Bani combined passes Mopti in dry years (13 km^{3} from the total flow of 20 km^{3}), but this fraction declines to 45% in wet years (30 km^{3} from the total flow of 65 km^{3}). The explanation is that a part of the water flows between Macina and Mopti directly into the Inner Niger Delta, primarily along the Diaka at Diafarabé, and further downstream, to a lesser degree, also along the Mayo Dipouli, Sogano, Mayo Moura, Mayo Kotaba and Mayo Dongue. The higher the river discharge, the larger the fraction flowing into the Inner Niger Delta upstream of Mopti.

*The total annual flow at incoming water measured in Mopti as a function of combined flow of the Niger (Macina) and the Bani (Bénény-Kegni) measured further upstream. The straight **line gives the expected relationship if both would be equal. At higher river discharge, an increasing fraction of the water does not reach Mopti, because more water from the Niger River flows after Macina, and before Mopti, directly into the Inner Niger Delta.*

## Annual inflow and peak water level

The annual peak flood level in the Inner Niger Delta, such as measured in Mopti, Akka and Diré, is determined by the flow of the Niger and the Bani. The three graphs show that the relationships are close, but not linear. Hence that third-degree polynomials were calculated to fit the relationships.

It is important to know these relationships and also that the relationships are so strong. Using these data, it is possible to indicate precisely what would be the impact on the flooding of an upstream infrastructure when it is known how much water is taken from the river during incoming water.

*The maximum annual flood level in Mopti as a function of the annual discharge of the Niger (measured at Macina) and the Bani (measured at Bénény-Kegni) during incoming water (1 June – 15 November).*

*The maximum annual flood level in Akka as a function of the annual discharge of the Niger (measured at Macina) and the Bani (measured at Bénény-Kegni) during incoming water (1 June – 15 November).*

*The maximum annual flood level in Diré as a function of the annual discharge of the Niger (measured at Macina) and the Bani (measured at Bénény-Kegni) during incoming water (1 June – 15 November).*

## Annual inflow and flood extent

A high inflow of the Bani and Niger Rivers not only produce a high flood level, but also floods a more extensive area. The higher the maximum flood level, the larger the area (see viewer). In a dry year (as 1984), just one third of the Delta became inundated; the northern Delta was not even reached by the flood. In a wet year (such as 1999) though, the southern Delta was fully flooded, as well as a large part of the northern Delta including several of the lakes just north of the Delta. Zwarts & Grigoras (2005) used a polynomial to describe the relationship between the peak flood level in Akka and the corresponding flood extent:

*maximal flood extent (km ^{2}) = 0.099x^{2} - 35.521x + 8727, where x = water level in Akka (cm).*

The function is based on water maps for which the water level in Akka was up to 511 cm, but in the past, peak flood level in Akka has been as high as 625 cm. Extrapolating this function for these higher water levels, the flood extent would arrive at 25 000 km^{2} when the water level in Akka would be 625 cm in Akka (as was the case in 1956). This is still substantially less than the total surface of the floodplains as shown on the IGN maps (31 000 km^{2}). This apparent discrepancy is caused by the shallow northward slope of the floodplain that delays flooding in the north with two-three months; by that time the southern floodplain has already been drained of water.

Using the function given above, the variation in flood extent can be shown for all years since 1907:

*The flood extent of the Inner Niger Delta derived from the annual peak flood level in Akka.*

Combining the data given above, it is also possible to show the relation between the total inflow of Bani and Niger Rivers and the flood extent in the same year:

*The maximal flood extent of the Inner Niger Delta as a function of the annual inflow of Niger and Bani during incoming water.*

The water maps given by Zwarts & Hoekema (2013) are intended to show the inundation at different water levels during the peak flood, but may also be used to show the change in the water cover during incoming water. The maps cannot be used, however, to show the water cover as a function of the water level during the deflooding when a large part of the inundated area lost its connection to the river system. The higher the flood, the more isolated lakes come into existence and the larger the fraction of inundated area being unconnected to the river. Hence, during the deflooding it is not the water level in the river, but the maximal water level, as well as the time passed since the water level has reached its peak, which determines where isolated and temporary lakes with water can be found. This is true for small depressions of some ha of shallow water, but also for Lac Faquibine and the other, large lakes west and east of the Inner Niger Delta.

## Seasonal variation in flooding

When the first rains arrive in May in the upper reaches of basin in Guinea, the water level in the Inner Niger Delta is nearly at its lowest. Then, the floodplain is desiccated apart from the river proper and some permanent water bodies. As rainfall progresses, the discharge of the Niger and Bani increases and gradually flooding starts in the most southern parts. The seasonal flood pulse enters the delta at Ké-Macina (Niger) and Sofara (Bani) and inundates a huge part of the floodplains between Djenné in the south and Tombouctou in the north.

The seasonal flood pulse is illustrated by the water levels at the stations in Mopti, Akka and Dire (see **water levels and river flows**). The flood moves as a wave through the delta, which means that the delta is never flooded entirely. Flooding starts in the south in August and in most years the peak in Mopti is reached in the first half of October. The water level in the centre and north of the delta is still rising at that time. In the centre of the delta the flood peak and maximal flooding in most years is reached in the second half of October or November, and in Dire about a week later. The southern floodplains already run dry at the time.

Also the annual variation in flood extent is large as a result of the varying discharge and rainfall. The figure below illustrates the large difference in flooding during the course of a dry and a wet year. In a dry year (as 1984), just one third of the Delta became inundated; the northern Delta was not even reached by the flood. In a wet year (such as 1999) though, the southern Delta was fully flooded, as well as a large part of the northern Delta including several of the lakes just north of the Delta.

*The graphs show daily measurements of the water level at Akka (left axis in both graphs) between June and May in 1984/85 (left graph; the lowest flood ever measured) and in 1999/2000 (right graph; one of the highest floods since 1970, but of normal height when compared with pre-1973 floods). The surface areas flooded (right axis in both graphs) in 1984/85 and 1999/2000, shaded in the same graph, are derived from the water level data, using the digital elevation models of Zwarts & Grigoras (2005). The maps show the flooded area when the water level at Akka reached its peak, and for 1 March. Note that the peak flood level in 1999 (511 cm) was 1.5 times higher than in 1984 (336 cm), but that the maximal flood extent was nearly 4 times larger. The difference is even larger some months later. The flood extent on 1 March (shown as dark blue on the map) was in 2000 13 times more than in 1985 (right map). From: Zwarts et al. (2009).*

## Sources and more information

- Brunet-Moret Y., Chaperon P., Lamagat JP. & Molinier M. 1986/ Monographie hydrologique du Fleuve Niger. Tome 1. Niger supérieur. Editions de l’ORSTOM ; Monographies Hydrologiques 8. 2605 pp.
- Davids, L., M. Bekkema, L. Zwarts, I. Grigoras 2018.

An improved spatial flooding model of the Inner Niger Delta. A&W-report 2529. Altenburg & Wymenga, Feanwâlden, Pays Bas. - Orange D, Arfi R, Kuper M, Morand P, Poncet Y, editors 2002. Gestion intégrée des ressources naturelles en zones inonables tropicales. Paris: IRD.
- Zwarts L. & Diallo M. 2002. Eco-hydrologie du Delta. In : Wymenga E., Kone B., van der Kamp J. & Zwarts L. editors. Delta intérieur du Fleuve Niger. Ecologie et gestion durable des ressources naturelles. Wetlands International,Altenburg & Wymenga, RIZA.
- Zwarts L, Cissé N, Diallo M. 2005a. Hydrology of the Upper Niger. In: Zwarts L, Beukering van P, Kone B, Wymenga E, editors. The Niger, a lifeline. Lelystad: RIZA/Wetlands International/IVM/A&W. p. 15-40. Mali-PIN-publication 2002-01.
- Zwarts L, van Beukering P., Kone B. & Wymenga E. 2005b. The Niger, a lifeline. Rijkswaterstaat/IVM/Wetlands International/A&W. p .1-304.
- Zwarts L. & Frerotte, J-L. 2012. Water crisis in the Inner Niger Delta (Mali). Causes, consequences, solutions.
- Zwarts, L & Hoekema, F.S. 2013. Atlas. Les plaines inondables du Delta Intérieur du Niger. Altenburg & Wymenga, ecological consultants, Feanwâlden.
- Zwarts L. & van der Kamp J. 2013. Does the Inner Niger Delta suffer from a reduced river flow in the dry season? A&W-rapport 1938. Altenburg & Wymenga ecologisch onderzoek, Feanwâlden.
- Zwarts L. 2012. L’impact d’un faible débit du fleuve sur l’inondation, la végétation et l’utilisation des terres dans le Delta Intérieur du Niger. A&W-report 1868 / 1877. Altenburg & Wymenga ecological consultants, Feanwâlden.