Significant baseflow reduction in the Sao Francisco River Basin

Water scarcity is a key challenge to global development. In Brazil, the Sao Francisco River Basin (SFB) has experienced water scarcity problems because of decreasing streamflow and increasing demands from multiple sectors. However, the drivers of decreased streamflow, particularly the potential role of surface-groundwater interaction, have not been yet investigated. Here, we assess long-term trends in baseflow, quickflow, and streamflow of the SFB during 1980– 2015 and constrain the most likely drivers of observed decreases through trend analysis of precipitation (P), evapotranspiration (ET), and terrestrial water storage change (TWS). We found that over 82% of the observed decrease in streamflow can be attributed to a significant decreasing baseflow trend (< -20 m3 s-1 y-1) along the SFR with spatial agreement between decreased baseflow, increased ET, and irrigated agricultural land. We also noted a decrease in TWS across the SFB with trends exceeding -20 mm y-1. Overall, our findings indicate that decreasing groundwater contributions (i.e., baseflow) is providing the observed reduction in total SFR flow. A lack of significant P trends indicates that only P variability likely has not caused the observed baseflow reduction, mainly in the Middle and Sub-middle SFB. Therefore, groundwater and surface withdrawals may be likely a driver of water scarcity over the SFB.


Introduction
The World Economic Forum has categorized water crises as one of the top-ranked global (severe) risks since 2012 [1], and water is increasingly becoming a priority policy issue at global scales [2]. Ensuring global access to freshwater is one of the United Nations Sustainable Development Goals (SDGs) addressed in the 2030 Agenda [3]. To fulfill the 2030 Agenda, decisionmakers face the complex challenge of balancing water resource availability and water demands [2,4]. Unfortunately, the imbalance between water supply and demand has led to water scarcity [5]. In fact, there is increasing trend in water scarcity worldwide [6] triggered in many locations by the increase anthropogenic water use [7].
Streamflow reduction is one manifestation of water scarcity [8]. Hence, understanding how human activities alter streamflow is vital to understand and manage future water scarcity [9,10]. Globally, long-term annual average streamflow has decreased by 2.7%, and monthly low flows have decreased, on average, by 57% on 26% of global land area because of increasing water abstractions and human water use [11]. Exacerbating these challenges, climate change (i.e., changes in precipitation and temperature) has also potential to alter river flow across the world [12][13][14][15].
Water scarcity can also manifest itself in the form of persistent groundwater depletion (or overexploitation) [16]. Persistent groundwater depletion has widely been reported in both regionally and [17][18][19][20][21][22][23][24], and quantifying groundwater depletion has been facilitated by the advent of the Terrestrial Water Storage (TWS) product measured by the Gravity Recovery and Climate Experiment (GRACE) satellite [25]. Additionally, groundwater withdrawals can decrease river flow where the stream and aquifer are hydraulically connected [26], a phenomenon known as streamflow depletion [27][28][29]. Globally, groundwater withdrawals are high and environmental streamflow limits are known to be severely exceeded in many locations [30], showing that groundwater use has increased water scarcity.
In Brazil, the Sao Francisco River Basin (SFB) has an national strategic importance due to its potential for agriculture, hydropower electricity, urban and industrial water supply and tourism [31]. The Sao Francisco River (SFR) is popularly called "the river of national integration" because crosses a variety of biomes, climates (including the semiarid region), landscapes and socioeconomic status throughout its extension, linking the southeast and northeast Brazil. The SFR is the fourth longest river in the Latin America. The SFB has faced serious water-related problems because of water conflicts for multiple uses and particularly importance for food production by irrigation. As the interaction between surface water and groundwater has often been neglected, the Brazilian government and press have attributed the decrease in streamflow solely to natural dry weather and/or droughts over the SFB. The premise behind this argument is that low precipitation conditions are the sole driver of decreasing flow in the SFR. However, the potential impact of groundwater withdrawals within the SFB on flow in the SFR has not been evaluated. This paper addresses for the first time some important issues pertaining to the SFR: we start by assessing whether there are significant decreases in streamflow in the SFR. After that, we investigate how baseflow are related to the observed SFR trends. Finally, we analyze the spatial and temporal patterns of climatic drivers and water storage at the SFR in order to understand the role of climate and groundwater withdrawals (demand driven) in explaining the observed trends in baseflow. Our findings represent a starting point to understand the impact of groundwater withdrawal over the SFB.

Study area description
Our study area, the SFB, is challenged by water conflicts for multiple uses, with irrigation for food production representing the largest. The SFB is located in northeast Brazil and covers 639,000 km² over seven Brazilian States: Bahia (contains 48.2% of the SFB), Minas Gerais (36.8%), Pernambuco (10.9%), Alagoas (2.2%), Sergipe (1.2%), Goias (0.5%) and Federal District (0.2%) (Figure 1). The SFR has a length of 2,600 km from its headwaters (northern Minas Gerais State) to estuary (between Alagoas and Sergipe States) in the Atlantic Ocean. The annual average flow and the 95th percentile flow (i.e., Q95 -a metric of low flows) of SFR is 2,914 m 3 s -1 and 875 m 3 s -1 , respectively [32].
Irrigated agriculture is the most important economic activity in the SFB and is responsible for 86% of permitted surface water withdrawals [33]. However, the SFR also supplies water for ~16 million people in 521 municipalities [34] and has hydroelectric power plants which supply ~12% (10,708 MW) of the installed generation capacity in the Brazil [35]. The SFB accounts for 23% (~750 m 3 s -1 ) of all the surface water abstractions permits in Brazil [33]. Although the average precipitation is 938 mm y -1 over the entire SFB [32], there is large variation in precipitation within the SFB (Error! Reference source not found.). Due to its variability in precipitation regimes and biomes, the SFB is divided into four hydrographical regions: Upper, Middle, Sub-middle, and Lower ( Figure 1). The main characteristics (e.g., precipitation, climate classification and biome) of each hydrographical region are summarized in Error! Reference source not found.. The entire Sub-middle SFB, and most of the Lower and Middle SFB, are in the Brazilian Semiarid climate zone (Figure 1), which is considered the driest region in the country due to prolonged droughts. The spatial distribution of precipitation, evapotranspiration and climate type are shown in Figure 2.  3 Aw and BShw (semiarid) 3 BShw 3 As (hot and humid with wet winter) and 3BSh (semiarid with short wet season) [39] 1 Cerrado is a Brazilian (tropical) savanna. 2 Caatinga is a Brazilian herbaceous and arborescent vegetation in the semiarid region. 3 According to Köppen's climate classification.

Precipitation, evapotranspiration and water storage change data
To evaluate potential drivers of streamflow changes in the SFB, we computed spatiotemporal trends in water fluxes (precipitation, P, evapotranspiration, ET, and total water storage, TWS). We used a ground-based gridded rainfall to retrieve daily P data (from 1980 to 2015) with 0.25° × 0.25° spatial resolution [40]. The gridded P data were obtained by interpolation technique using data from ~4,000 rain gauges over the Brazil [i.e., 40] and have been widely used in regional studies [41][42][43][44].
Daily ET and potential ET (PET) was acquired from the Global Land Evaporation Amsterdam Model (GLEAM) dataset [45], with 0.25° × 0.25° spatial resolution during the 1980-2015 period, which is available at https://www.gleam.eu/. GLEAM is dedicated to the estimation of terrestrial evaporation and root-zone soil moisture from satellite data [46]. Basically, GLEAM separately derives the different components of terrestrial evaporation, i.e. transpiration (from short and tall vegetation), bare soil evaporation, open-water evaporation, interception loss, and sublimation on daily basis [45]. GLEAM has been used to evaluate trends in ET at global scale [47].
Additionally, we analyzed monthly TWS anomaly with 1° × 1° spatial resolution (from mid-2002 to 2015) measured by the Gravity Recovery and Climate Experiment (GRACE) satellite mission. GRACE measures temporal variations in the Earth's gravity field, which can be used to estimate changes in TWS (i.e., surface water, snow, groundwater and soil moisture storage) [48]. GRACE-based monthly gravity products are officially processed and distributed by three processing centers: GeoforschungsZentrum Potsdam (GFZ), Jet Propulsion Laboratory (JPL) and Center for Space Research of the University of Texas at Austin (CSR). Traditionally, two approaches have been used to generate GRACE-based TWS solutions: spherical harmonics and mass concentration blocks (mascons) [49]. Here, we used JPL Release 06 (RL06) mascons solution, which is available at https://grace.jpl.nasa.gov/data/get-data/jpl_global_mascons/.
Trends in P, ET, and TWS dataset were analyzed on annual and seasonal basis. Seasons were defined as: DJF (December, January and February), MAM (March, April and May), JJA (June, July and August) and SON (September, October and November). We used the location of groundwater wells and information about irrigation withdrawals from the Brazilian Geological Survey (CPRM). Further, we acquired land use and land cover information over the SFB from MapBiomas Brazil Project. MapBiomas is a collaborative multi-institutional initiative to generate annual land cover and use maps in 30 m resolution using automatic classification processes applied to satellite images [50]. The complete description of the MapBiomas Brazil Project and dataset can be accessed at http://mapbiomas.org.

Partitioning of streamflow into baseflow and quickflow
We quantified trends in total streamflow (Qt), baseflow (Qb), representing slowly varying inputs to streamflow such as groundwater), and quickflow (Qq), representing quickly varying inputs to streamflow such as surface runoff). The baseflow regime was taken as representative of the overall behavior of groundwater contribution over the SFB. Digital filters have been widely used to estimate the contribution of groundwater to streamflow [42,51] and have been shown suitable in comparison to other approaches over large spatial scales [52][53][54]. We applied two digital filters to decompose Qt into Qb and Qq components at daily resolution, and used the average value of Qb of the two filters. The Qt, Qb and Qq at daily resolution was then aggregated at the annual timescale.
First, the one-parameter recursive digital filter [55] was applied to estimate Qb on daily time step (i) as follows: where ϵ is the filter parameter which should assume a single value of 0.925 to generate realistic comparison to manual decomposition techniques [51,56].
We also applied a second, two-parameter digital filter by Eckhardt [57] at a daily time step given by: where BFI max is the maximum value of the baseflow index (Eq. 4) that can be modelled by Eq. (3) and a is a constant (see Eckhardt, 2008). Following Eckhardt [57], we adopted BFI max = 0.80 and a = 0.98 (i.e., perennial streams with porous aquifers). These values are specifically suitable for some important (granular) porous hydrogeological units presented in the SBF, for example, the Urucuia Group, which represents an important source of groundwater contribution to SFR flow [59].
To evaluate the role of groundwater contribution to SFR, we chose baseflow index (BFI) as hydrological signature. The BFI represents how extensive groundwater contribution is to streamflow [60]. It was calculated as a long-term ratio between Q b and Q t (Eq. 4) at daily time step.
The BFI was calculated for each gauge. High BFI values (> 0.70) indicate that groundwater is the major contributor on streamflow [61]. Conversely, low BFI values (BFI < 0.40) indicate that quickflow is predominant in terms of streamflow, attesting greater surface runoff [61].    The red arrow represents significant decreasing trend. BFI = baseflow index. -means nonsignificant trend.

Trends in streamflow, baseflow and quickflow
In general, we found a significant decreasing trend in Q t , Q b , and Q q (at α = 0.05) over the SFB during 1980-2015 (Error! Reference source not found.). Trends in the two gauges located in the Upper SFB were not significant, while all other gauges had significant decreasing trends for all components of flow. Averaged over all gauges, Q b in the SFB is decreasing at a rate of -26. became significantly more negative. Overall, the trends observed in Q b explain the large majority (82%) of the observed trends in Q t , indicating that observed decreases in streamflow can be primarily be attributed to changing baseflow conditions. Similar to trends in Q t and Q b , significant decreases in Q q were found throughout the SFB (Error! Reference source not found.). Trend magnitude in Q q ranged from -3.19 to -5.69 m 3 s -1 y -1 in the Middle, from -5.71 to 6.55 m 3 s -1 y -1 in the Sub-middle, and between -5.65 and -7.23 m 3 s -1 y -1 in the Lower SFB. Overall, trend magnitude in Q q became more negative from the Middle to Sub-middle and Lower SFB. Our spatial analysis approach did not explain the significant decreases in Q q .

Trends in precipitation, evapotranspiration and water storage changes
The annual P and ET were not significant (at α = 0.05) over most of the SFB during 1980-2015 (Figure 3c,d). Nevertheless, a significant decreasing trend in annual P was observed in few regions over the western Middle SFB, specifically next to the western edge of the basin (Figure 3c) which is a region with widespread irrigation (Figure 3b). In addition, annual ET tended to decrease in few regions over central and western part of the Middle SFB and in the Upper SFB (Figure 3d). We also noted a significant negative trend in TWS over almost the SFB, mainly over Middle SFB (Figure 3e). On the other hand, a nonsignificant trend in TWS was noted in the Lower SFB (Figure 3e). One should note the overlap between regions with decreasing P and TWS, decreasing Q b , and increasing ET in the Middle SFB, which has most irrigated agriculture in the region (Figure 3b-e). A strong trend magnitude in TWS of -2.3 mm y -1 occurred in the Middle SFB around the SFR channel path near regions with large irrigation withdrawals (Figure 3b,e). Furthermore, we noted some regions with a significant decreasing trend in TWS (-1.4 ≤ TWS ≤ -0.5 mm y -1 ; Figure 3e) over the Sub-middle SFB, which has a large number of irrigation wells (Figure 3a). However, information about the water demand and purpose of individual wells (e.g., irrigation and urban supply system) were not available. Moreover, although these wells are documented we cannot identify which ones were in operation.
The seasonal analysis revealed an overall nonsignificant trend in P over the SFB (Error !  Reference source not found.a-d). The exception is a strong decreasing in P (-12 ≤ P ≤ -1.0 mm y -1 ) over the western Middle SFB during DJF season (Error! Reference source not found.a). In addition, the P also tended to increase (P = 9 mm y -1 ) strongly in a few regions over and adjacent to the Lower SFB during JJA season (Error! Reference source not found.c). Thus, in general, the seasonal P results indicate that the decreases in annual P are primarily driven by reductions in summer (DJF) precipitation. With regard to TWS, we noted a negative trend in the western and eastern Middle SFB during all seasons (Error! Reference source not found.i-l) where there are large irrigation demand and wells (Figure 3a,b). For instance, the strongest seasonal decreasing trend (-40 ≤ TWS ≤ -20 mm y -1 ) overlapped the agricultural irrigation land over the western and central Middle SFB in MAM and JJA (Error! Reference source not found.j,k). Furthermore, TWS significantly increased from JJA to SON and DJF season in the Centre-West and Centre-East of Middle SFB. Conversely, there was a nonsignificant seasonal trend in the Upper, Sub-middle and Lower SFB.

Discussion
The partitioning of total streamflow into baseflow and quickflow components reveals the major role of groundwater in sustaining the Sao Francisco River discharge. Our results show a strong decreasing trend in baseflow over almost the entire SFB (Error! Reference source not found.), most prominently in the heavily-irrigated Middle SFB, and that baseflow reductions are the primary driver of observed change in total streamflow. The Upper SFB is the only region that presented nonsignificant trends in baseflow during the 1980-2015 period. These results suggest that groundwater withdrawals for irrigation may be the most likely driver responsible for decreased baseflow in the Middle SFB and downstream areas. This is supported by the spatial correspondence between irrigated agricultural land (Figure 3b), a decreasing trend in TWS (Error! Reference source not found.i-l), and a strong increasing trend in ET (Error! Reference source not found.f,g). While there are many irrigation wells in this region (Figure 3a), we do not have data on direct surface water withdrawals from the SFR, which may also contribute to strong decreasing trend in Q b near the irrigated land (Figure 3e).
The close relation between increased ET and decreased P over the irrigated area (i.e., Centre-West part of the Middle SFB, which comprises the MATOPIBA region) can be related to deforestation for agricultural cropping. MATOPIBA is located mainly in the Bahia and Tocantins States with irrigated holdings larger than 5,000 ha and water demand is greater than 100 L s -1 [62]. The observed decrease in ET during SON season may be explained by the harvesting of different crops, mainly maize (known as "safrinha"), sugarcane and wheat [66]. This raises the possibility that the harvest of crops may trigger a deforestation-induced feedback phenomena in which reduced ET can further reduce P, as has been demonstrated in the Amazon forest [i.e., 66]. This is corroborated by the increased TWS from MAM and JJA to SON and DJF period, indicating more infiltration, and eventually groundwater recharge. These results agreed with those reported by [62,68]. Although this area also comprises rain-fed agriculture, irrigation is often required during dry periods. Unfortunately, data are not available on the seasonal distribution of groundwater and surface water withdrawals.  The decrease in baseflow can be partly related to groundwater irrigation from the Urucuia Aquifer System (UAS). The UAS is an sedimentary aquifer that covers an area of 125,000 km 2 and is an important source of water to the SFR, especially during dry seasons [69]. The UAS is located in most part over the western Bahia State by the left bank of SFR and covers the MATOPIBA region. On average, the UAS discharged 792 m 3 s -1 to the SFR in 1980, however this contribution has decreased continuously to 390 m 3 s -1 in 2015 [70]. In addition, previous studies showed that groundwater storage tended to decrease in -6.5 ± 2.6 mm y -1 (9.75 km 3 ) in the UAS over the 2003-2014 period [69]. Therefore, given the importance of the UAS, it is reasonable to infer that groundwater decline in the UAS may have affected the SFR flow, particularly in the Middle SFB, during the 1980-2015 period.
In the Sub-middle SFB, it is likely that baseflow reduction is driven by combined surface and groundwater withdrawals. Annual decreasing trend in TWS occurred only in the Centre-North part of Sub-middle (near the SFR) and was lower related to the Middle SFB. This indicated that the surface water withdrawals for irrigation may be responsible for the lack of TWS trends in the regions south of the SFR in the Sub-middle SFB. The Sub-middle SFB has been subjected to expansion of irrigated agriculture (Figure 4), primarily vineyards and mango orchards in the Juazeiro (Bahia State) and Petrolina city (Pernambuco State) [71]. To carry out the irrigated expansion in this region, the undisturbed vegetation (Caatinga) have been replaced by vineyards [72]. Since mango orchards and vineyards are perennial fruit crops, irrigation from surface and/or groundwater is needed during dry periods. [73] reported the impact of replacing Caatinga to vineyards and mango orchard on ET over the Sub-middle SFB. They concluded that ET increased by 2 and 3 times for vineyards (450-800 mm y -1 ) and mango orchard (500-1,300 mm y -1 ), respectively, in comparison with the Caatinga vegetation (200-400 mm y -1 ) in 2006 [73]. However, we surprisingly found a nonsignificant trend in ET over the Sub-middle SFB (Figure 3,Error! Reference source not found.). This may occurred due to the longterm (average) relation between ET, potential evapotranspiration (PET) and P. Using the Budyko framework [74], we find that ET is water-limited by the dry climate (i.e., high temperature and solar radiation) over the Sub-middle SFB ( Figure 5). As consequence, significant trends in ET were not detected even with the increased ET after replacing Caatinga vegetation by vineyards and mango orchard. Therefore, it is reasonable to infer that the climatic condition (i.e., P < ET < PET) may be another responsible to the decrease trend in TWS in these regions. However, according to Figure 5 other factors like surface and groundwater withdrawals may explain observed departures from observed streamflow against the Budyko curve. The streamflow depletion we describe above causes ET to be overestimated when calculated as Q t minus P following the Budyko framework, leading to the points shifted upwards above the Budyko curve ( Figure 5). Further, climate variability (i.e., P and ET ) may overlap the effect of groundwater withdrawals during the study period, making it difficult to evaluate them separately.

Conclusion
Decreasing streamflow has brought a major concern about water security worldwide, especially in (semi)arid regions. In Brazil, the Sao Francisco Basin (SFB) has experienced water-related problems because of decreased flow in Sao Francisco River (SFR). The SFB is strategically important for national food and energy production with a huge irrigated area, and a population of ~16 million.
These multiple water demands make the SFR a socio-economically important region and balancing Preprints (www.preprints.org) | NOT PEER-REVIEWED | Posted: 17 September 2020 doi:10.20944/preprints202009.0410.v1 their needs poses a difficult task for decisionmakers. Because the Brazilian government has encouraged the expansion of irrigated agriculture to increase food production, we further expect increased water demand and conflict in the future of the SFB [75].
Although previous studies have focused on the impacts of climate variability on decreasing SFR flow, none has addressed the possible influence of groundwater withdrawals. Here, we provide the first overview of long-term change in relevant hydrological fluxes (i.e., precipitation, evapotranspiration, baseflow and water storage) to constrain potential drivers of the observed decrease in SFR flow. Our findings suggest that significant decreases in SFR flow are primarily due to reduced groundwater contributions to total streamflow (as baseflow) during the 1980-2015 period.
While no long-term groundwater use data are available for the region, this suggests that groundwater withdrawals may be the most likely driver of the observed negative trend in the SFR baseflow, which is compounded by direct surface water withdrawals. Further investigation is necessary to precisely identify the causes of decreased baseflow and relative importance of different drivers.
Combined, these results indicate that river water and groundwater are hydraulically connected in the SFB, and that groundwater pumping can decrease flow in the SFR. Therefore, we call for increased attention on groundwater management in the SFB, especially in the most heavilyirrigated Middle SFB region. As irrigated agricultural will likely continue to expand and require more water, groundwater will play a key role for future food production.