Purposeful destruction of civilian water infrastructure is a violation of international law (UN, 1997) (International Law Association, 2004). However, water interruption during armed conflict may also occur indirectly as a result of the use of heavy weapons which can cause indirect but significant damage to essential infrastructure. Such ‘explosive weapons with wide area effects’ as termed by the International Committee of the Red Cross (ICRC) have a myriad of effects on civilians and disruption to services including electricity, WASH and healthcare. (ICRC, 2022) The long-term health impacts of such damage are harder to quantify but the use of heavy munitions is increasing in modern armed conflicts.
From our literature review, we noted 14 disruption events in Ar-Raqqa, Deir-ez-Zor and Al-Hasakeh cities, which were the most populous cities in the region and have been the focal point of the conflict to date. Similar events and findings were reported by Abbara et al. in their investigation into the weaponization of water in northwest Syria in Aleppo and Idlib governorates, with the majority of events occurring in the main cities (Abbara et al., 2021).
Interference with water also occurs elsewhere in the MENA region such as in Palestine where pumping rates and water consumption differ significantly between Israel and Palestine despite the two nations despite sharing aquifers, with evidence noted that Israel regularly violated the limits set out in the Oslo Agreement and that the countries geographical advantage in access to water allowed it to exploit Palestine (Zeitoun et al., 2009). As a result, a framework of “hydro-hegemony” was conceptualised, which addresses the role that power asymmetry plays in creating and maintaining transnational conflict over shared water sources (Zeitoun and Warner, 2006). Spiegel et al’s investigation into cholera in Yemen also noted the impacts of insecurity (specifically, airstrikes) and conflict on preparedness planning and community-directed responses. (Spiegel, 2019) For NES, interference with WASH will have an increasingly important influence due to water sources and increasing water shortages.
Disease and surveillance data
That 80% of disease reports were “other acute diarrhoea” deserves some attention to ensure accurate reporting of the case definitions for acute bloody and acute watery diarrhoea by health workers. An overall upwards trend of WBDs, with seasonal peaks in the late summer, was observed throughout the whole region, corresponding to trends reported in northwest Syria (Abbara et al., 2021). However, considerable variability in trends and median weekly levels of WBDs were observed in different districts. Trends of waterborne diseases in areas of conflict are likely to be affected by various factors such as conflict, reporting accuracy and population movement (Schillinger et al., 2020). Establishing cause and effect is geo-temporally complex. For example, Ain al Arab district consistently reported some of the lowest median incidence of WBDs, despite having relatively high completeness and timeliness throughout the study period (data not published), indicating potential differences in detection and reporting capacity. A previous study using survey data on WASH access in southern Syria found that a significant risk factor for the incidence of childhood diarrhoea was whether the household shelter of the responder was damaged. Significant protective factors were found to be hygiene access and reported handwashing within the household (Sikder et al., 2018). Therefore, it is likely that factors including hygiene access and living conditions may be important factors in the incidence of WBD.
The risk of cholera is also greatly increased in areas of conflict; in Yemen, targeting of hospitals, clinics and water infrastructure led to the largest cholera outbreak in the world (Spiegel et al., 2019). Despite concerns about the potential for cholera to spread in Syria due to the poor water and sanitation levels, outbreaks in neighbouring Iraq, and unreliable disease surveillance (Sparrow et al., 2016), AWD (indicative of cholera) had very low case levels in NES. There was one alert of an RDT-positive case in 2015 by EWARN for which no culture could be obtained due to ongoing conflict. Here, the lack of outbreak detection may be explained by a failure to initiate regular investigation and response protocols due to lack of access (Ratnayake et al., 2020). However, in our analysis, only 8 cases were reported across the entire study period and in two alert-level events in Raqqa. This may represent sporadic cases without further transmission or a failure to initiate regular investigation and response due to limited access or resources (Ratnayake, 2020)
Relationship between disruptions to water infrastructure and waterborne diseases
The relationship between weaponization of water and waterborne diseases has been investigated in a small number of reports. In northwest Syria, the peak of disease incidence occurred during an intensification of attacks on water infrastructure, although no direct correlation could be drawn (Abbara et al., 2021). A broader relationship between conflict, quantified by records of improvised explosive devices, and incidence of polio in Afghanistan has also been reported. The association was thought to be predominantly due to disruption to childhood polio vaccination due to the conflict. However, causation could again not be established due to data constraints (Norris et al., 2016).
Our hypothesis is that disruptions to water infrastructure could lead to increased incidence of WBDs, perhaps extending to months or a year after the disruption. This appears to be a possibility for Alouk water station which had repeated and prolonged interruptions. For the other 13 disruption events noted in the literature, there was no clear evidence of short- or longer-term effects and therefore we did not attempt an interrupted time series analysis. This may be for a number of reasons including insufficient information in the literature about the extent and duration of disruptions to water during the study period which speaks to the need for systematic and more accurate conflict mapping; that investigated WBDs to district level gives insufficient granularity thereby missing effects at the subdistrict level; decreased detection and reporting of the EWARN system particularly during escalations of conflict; the use of mitigating factors e.g. water trucking which averted WBDs outbreaks; the difficult in isolating the population impacted for a water system which may extend to several sub-districts. Additionally, it may be that the repeatedly affected population may already be chronically water insecure such that the impact of further, acute disruption could not be elicited.
Limitations
An important limitation in this study is the lack of credible, consistent reports of attacks against WASH. Unlike the case for Yemen, where the Yemen Data Project (https://yemendataproject.org/) systematically compiles verified lists of air raids, no similar efforts are in place in Syria (Spiegel, 2019). Despite triangulation to verify sources where possible, the majority of sources used were grey literature or news articles, few published accounts; without a systematic process, we suspect a high degree of undercapture of events. For WBDs, there may have been under-reporting into EWARN, especially during periods of violent attacks. Our data analysis went to the district level, potentially missing changes at the subdistrict level. It is also known that many in NES, particularly in camp settings, rely on trucked water from either private companies or by humanitarian actors (UNICEF, 2017). These may therefore mitigate the effects of interruptions to WASH in the area and are harder to account for. Lastly, our population denominators are estimates given the lack of one regional civil record with accurate numbers in this context.