Climate Variability and Water Availability in Ambositra, Amoron’i Mania Madagascar

This work provides a statistical analysis of urban water shortages, as well as the direct and indirect impacts of the scarcity of rainfall on agricultural practices in the Isaha watershed, Amoron’i Mania Region Madagascar. 78% of the urban population no longer has normal access to drinking water, and 85% of farmers report difficulties with their practices. Paradoxically, according to WEAP modelling, there is sufficient water available. Because of poorly management and other factors considerable loss of water is observed.

Introduction

In Madagascar, the year 2021 was marked by a shortfall in rainfall compared with the average, particularly during the hot periods that correspond to active vegetative periods [1]. The Amoron’i Mania Region was not excluded from this situation, with a negative SPI (Standardized Precipitation Index) this year [1]. This index is used to measure meteorological drought, which is based on precipitation. Probabilities are standardized, so that an SPI of 0 indicates a median precipitation quantity (compared with a reference average climatology, calculated over 30 years). The index is negative for drought and positive for humid conditions [2]. This situation has been severely felt by a vulnerable urban and rural population. The allocation of limited water resources, environmental quality, and sustained water use policies are issues of growing interest [3]. This led to the choice of this research topic, whose main objective is to assess the direct effects of rainfall deficit on anthropogenic activities in the Isaha, Ambositra watershed. On a watershed scale, the study area covers the entire section bordering the Isaha river and includes several categories of area. The activities therefore, concern both the water needs of the urban population and the water needs of agriculture in urban, suburban, and rural areas.

Materials and Methods

This section attempts to provide a more detailed framework for the study area, with the necessary data to serve as a basis for analysis and the scientific approaches used to carry out this work.

Presentation of the Study Area

The Isaha watershed, shown as a water body in Fig. 1, is administratively located in the rural Communes of Ankazoambo, Ambositra II, Ivato, and the southern part of Tsarasaotra, in the urban Commune of Ambositra I; in the Ambositra District, Amoron’i Mania Region. The basic hydrology of the Amoron’i Mania Region is dominated by the Tsiribihina catchment area. The hydrological network of this catchment area originates in the region and empties into the Mozambique Channel, once recovered by the Tsiribihina River [4]. The main rivers are the Mania and its tributaries (Izanaka, Ikely, Imorona and Ivato). The Isaha river is a permanent body of water originating from these streams [5].

Fig. 1. The study area with Hydrographic network (Author, 2021).

Prior to the town’s creation, a now-extinct plant species vernacularly named “Somangana” dominated these riparian habitats [6], which was the origin of the older, indigenous name Sahasomangana, currently Isaha; with Sahasomangana iray on the southern part of the town (iray to say upstream) and Sahasomangana iava on the northern part of the town (iava to say downstream)¹1This shows how important the Isaha, a tributary of the Mania, was to the local population., hence the origin of the word Isaha²2Diminutive of the word Sahasomangana, the I is placed before proper nouns for the general appellation in Malagasy., the subject of this study. Ambositra is 255 km from Madagascar’s capital via the RN7 to the south.

The climate of this area, like that of the entire Amoron’i Mania Region, is characterized by a cool, dry season (ririnina) and a warm, wet season (fahavaratra), from October to April, although current climate change is leading to variability in seasonal cycles and climatic parameters. According to the Köppen-Geiger Climate classification for the period 1991–2020, this Region is subject to a subtropical highland climate [7]. Precipitation in Madagascar is influenced by easterly winds (Alizé), the temperature of the Indian Ocean, tropical convergence zones, the Foehn effect, and the El Niño phenomenon [8].

Methodology

The approaches adopted for this work combine four interdependent stages: (a) acquisition of rainfall data, (b) calculation of SPI (Standardized Precipitation Index), (c) modeling of water resource use using WEAP (Water Evaluation And Planning) software (d) a survey of the population for a very practical overview of the water availability situation in the city.

Precipitation Data Acquisition

Historical precipitation data are obtained from the World Bank’s Climate Change Knowledge Portal³3CCKP-WBG (Climate Change Knowledge Portal of the World Bank Group)  https://climateknowledgeportal.worldbank.org/.. This portal provides historical and future data from climate modelling based on the RCP 4.5 scenario.

Calculation of SPI (Standardized Precipitation Index)

The SPI [2] is a powerful, flexible, and easy-to-calculate index. Indeed, it is equally effective for water use. What’s more, it’s just as effective for analyzing wet periods/cycles as it is for analyzing dry periods/cycles. Ideally, at least 20 to 30 years of monthly values should be available [9]. Hence, there is a need for the above-mentioned historical precipitation data.

The index is negative for drought and positive for wet conditions [2].

The SPI (Table I) has been designed to measure precipitation deficit on several time scales. These time scales reflect the impact of drought on the availability of different water resources. Soil moisture Soil moisture conditions respond to precipitation anomalies on a relatively short time scale. Groundwater, river flow, and reservoir storage reflect longer-term precipitation anomalies.

Table I. SPI Values [2]

2.0+	Extremely humid
1.5 to 1.99	Very damp
1.0 to 1.49	Moderately humid
−0.99 to 0.99	Close to normal
−1.0 to −1.49	Medium-dry
−1.5 to −1.99	Severely dry
−2 and below	Extremely dry

The formula used is:

SPIijk=Pijk−Pijσijwhere SPIijk is the z-value for study area i during period j for year k, Pijk is the precipitation value for study area i during period j for year k, Pij is the mean for study area i during period j over n years, and σij is the standard deviation for study area i during period j over n years [10].

Let’s take a concrete example (January 2000):

• SPIijk: SPI for the study area for January 2000.
• Pijk: Precipitation value for the study area for January 2000.
• Pij: Average January precipitation for the study area over 30 years.
• σij: Standard deviation of January precipitation over the 30 years and so on for the other timeframes.

Modeling Water Resource Use with WEAP Software

The basic data on the use of water resources at Isaha are based on the results of modelling with WEAP (Water Evaluation and Planning) software from Stockholm University. This part is based on the method adopted by Randriamifidison&al in 2021. WEAP is an integrated water resources planning tool for microcomputers. It provides a comprehensive, flexible, and easy-to-use structure.

The various modeling steps in WEAP are as follows:

• Study area creation: In vector form.
• Setting general parameters, including period,
• Creation of request sites: After digitizing the region’s main hydrographic network, four request sites (population, agriculture) were created.
• Creation of key assumptions: Two assumptions were adopted: the population’s need for water varies with population growth, and climate variability affects the availability of resources for population and agricultural needs.
• Scenario creation: These are variables constructed from key assumptions based on the current state,
• For the “Climate change” scenario, the weather patterns (very dry, dry, normal, wet, very wet) were assigned numerical values (Table II) so that the software could recognize them.

Table II. Numerical Values for Each Climate Type [5]

Climate types	Corresponding values
Very dry	0.7
Dry	0.8
Normal	1
Wet	1.3
Very damp	1.45

The data obtained with WEAP are then analyzed using data from user surveys, of which 150 were randomly distributed in Fokontany.

Experimental Approach to People’s Access to Water Resources, Based on Field Practices

Interviews with the JIRAMA (national water and electricity company) sub-group and closed surveys of the Ambositra population were conducted. A map was drawn up to give a better idea of access to water in the town during the shortage period. The survey sites covered twenty of the twenty-two Fokontany (the smallest administrative division in Madagascar) of Ambositra, mixing various parameters: center, periphery, administrative zone, residential, popular, commercial, populated, sparsely populated, etc. Two hundred households were visited. Two hundred households were visited, 63% of whom were heads of household. The category of people surveyed encompasses a wide range of jobs, both private and public, from students, civil servants, craftsmen, carpenters, grocers, agricultural workers, etc.

Results

This section provides an overview of the current situation, followed by appropriate recommendations.

Historical Precipitation Data

Since calculating the SPI requires 20 to 30 years of data, the data used for this work covers the period from 1991 to 2021, i.e., 30 continuous years. Data are missing for the year 2017, so considering values from 2018 onwards could distort the SPI obtained, as a break may change the value.

The red trend line shows a visible decline in precipitation (Fig. 2). Some “dry” years record below-average precipitation of 1204.64 mm (shown by the yellow dashed line).

Fig. 2. Annual precipitation for the Isaha watershed 1991–2021 [11].

SPI Annual Values

The 30-year precipitation data are used for the SPI calculation. The monthly and annual results are shown in Fig. 3 below. According to this figure, most years have an SPI below 0, i.e., negative. Remember that the index is negative for drought conditions and positive for wet conditions [2]. Since 2004, there have been no years with a positive SPI, which is nevertheless close to normal.

Fig. 3. SPI values for the last 30 years (1991–2021) for the Isaha River watershed.

Water Resources

Annual drinking water production by JIRAMA⁴4Société de distribution d’eau et d’électricité à Madagascar, these data are not published. is estimated at an average of 68823 m³, of which 65088 m³ is used (JIRAMA, 2019), with 2031 subscribers. Compared with the population, this figure is too low, with less than 3% of the population subscribing to water from JIRAMA, the sole distributor.

The following figures show the distribution of the population by type of resource (Figs. 4 and 5).

Fig. 4. Population distribution by resource type: (a) urban Ambositra, and (b) rural Ambositra (Author’s survey, 2022).

Fig. 5. Unequal distribution of water resources in the city.

Table III shows that the quantification of water resources in relation to surface water and groundwater gives a satisfactory result, taking into account the needs of the population and agriculture. In other words, water is sufficient according to this model. On the other hand, in 2021, the population of Ambositra encountered widespread difficulty in accessing water. This year, 78% of the urban population no longer had normal access to drinking water. Consumption increased by around 121087 m³ from 2017 to 2020. But this consumption falls sharply by 97385 m³ of water between 2020 and 2021 [8]. 85% of farmers declare having had difficulties with their practices according to our surveys. This is a striking contradiction.

Table III. Water Use Modeling for Isaha (Ambositra Urban, Ambositra Rural, and Outskirts) (WEAP Model)

	Jan	Feb	Mar	Apr	May	Jun	Jul	Aug	Sept	Oct	Nov	Dec	Sum
Area inflow
Below Isaha headflow	29.38	26.47	29.73	16.3	9.91	6.74	6.43	5.92	4.38	4.18	8.09	22.42	169.95
Area outflow
Farming demand	0.00	0.1	0.18	0.2	0	0	1.1	1.97	1.11	1.19	1.63	1.3	8.78
Population demand	0.00	0.00	0.00	0.00	0.00	0.00	0.00	0.00	0.00	0.00	0.00	0.00	0
Return from agricultural use	29.38	26.37	29.55	16.1	9.91	6.74	5.33	3.95	3.27	2.99	6.46	21.12	161.17
Sum	29.38	26.47	29.73	16.30	9.91	6.74	6.43	5.92	4.38	4.18	8.09	22.42	169.95
Unsatisfied demand
Farming demand	0.00	0.00	0.00	0.00	0.00	0.00	0.00	0.00	0.00	0.00	0.00	0.00	0.00
Population demand	0.00	0.00	0.00	0.00	0.00	0.00	0.00	0.00	0.00	0.00	0.00	0.00	0.00
Sum	0.00	0.00	0.00	0.00	0.00	0.00	0.00	0.00	0.00	0.00	0.00	0.00	0.00
Distribution requirement (with losses, recycling) Millions of cubic meters
Farming demand	0.00	0.17	0.30	0.33	0.00	0.00	1.83	3.29	1.85	1.98	2.71	2.17	14.63
Population demand	0.00	0.00	0.00	0.00	0.00	0.00	0.00	0.00	0.00	0.00	0.00	0.00	0.00
Sum	0.00	0.17	0.30	0.33	0.00	0.00	1.83	3.29	1.85	1.98	2.71	2.17	14.63

Urban Water Accessibility

Obvious disparities have been observed in terms of access to water within each Fokontany. In order to provide a better representation of water outage periods in each Fokontany (see Fig. 5), mapping was conducted. The blue area represents the portion where there were no water outages throughout the year. These correspond to the administrative site (Antaponivinany) and the low-lying areas (Ankorombe, Tanamahalana). For a better understanding, detailed information is provided in Fig. 6 for each neighborhood, including the average duration of outages.

Fig. 6. Access to water within each Fokontany. Source: authors’ surveys, April 2022.

Conclusion

At present, several regions in Madagascar are facing formidable freshwater management challenges. While the southern region used to suffer the most, today, all of Madagascar’s cities have water access problems. Admittedly, this may be due to technical shortcomings, but if the results of this research are anything to go by, it’s poor resource management. The situation is alarming because the calculation of SPI shows us years of drought becoming increasingly severe. Despite satisfactory modelling with WEAP software, the reality is that infiltration is very limited with the disappearance of vegetation cover. And the Integrated Water Resources Management (IWRM) approach is not being properly applied. We also need to disassociate the theory of water availability equals to rainfall availability. Time is of the essence, and strategies should be adopted based on the data put forward by this work. Combat surface water run-off during wet periods, adapt the habits and practices of the population and adapt the agricultural production system to the variability and risks of drought. Sound the alarm for drought intensification with SPI results, which should then be analyzed with climate projection data to anticipate problems and prepare for them. SPI values are close to normal, but poor management is hampering long-term access to water. Added to this are the phenomena of rapid evaporation of available resources without appropriate measures and storage facilities.