Seasonality and criteria for concession of water in the Ivinhema basin

Global demand for water has been increasing per year due to population growth, economic development, and changes in consumption patterns, among other factors. This increase in water demand is expected to continue in the next decades. The objective of this work was to evaluate the use of different criteria to grant the use of water from the Ivinhema river basin, Brazil. Monthly periods were compared to annual periods to calculate the reference flows Q7,10 and Q95. The relative differences in water availability using different reference flow rates for water concession were quantified. The replacement of the annual criteria (standard in Brazil) for water concession by 50% of monthly Q7,10 and 70% of monthly Q95 can potentially Research, Society and Development, v. 9, n. 10, e1969108391, 2020 (CC BY 4.0) | ISSN 2525-3409 | DOI: http://dx.doi.org/10.33448/rsd-v9i10.8391 2 increase the use and improve the management of water resources. The best criteria to award grants is the monthly Q7,10, which despite being more restrictive, it allows higher flow rates when there is excess water, and lower rates in the months of low water availability.


Introduction
Global demand for water has been increasing approximately 1% per year (WWAP, 2012) due to population growth, economic development, and changes in consumption patterns, among other factors. This increase in water demand is expected to continue in the next decades (Unesco, 2018).
In the context of water resources management, minimum flows have received attention because they represent the conditions of a basin during the dry season, when the water supply is limited (Silva et al., 2017). For an adequate management of shared water resources, it is essential to precisely estimate the water flows (Li et al., 2010;Arai et al., 2012;Pinto et al., 2015). Commonly, the average minimum flow of seven consecutive days with return period of 10 years (Q7,10) and the minimum expected flow in 90% or 95% of the time (Q90 or Q95, respectively) are adopted as reference (Brodie et al., 2008;Tucci, 2012).
The procedure for water concession is based on the minimum reference flow, which corresponds to the conditions during the greatest water shortage in a given year. This annual value can restrict the water use. However, where the water demand is high, a larger volume of water could be granted in the concession, especially during rainy periods. Therefore, it is necessary to consider the seasonal nature for the granting criteria.
There is great water availability in Mato Grosso do Sul, Brazil. However, the management of water resources requires solutions that satisfy the increasing challenges of water security arising from population growth and climate change. The water catchment area of Ivinhema is the second largest basin in Mato Grosso do Sul. It provides water to approximately 26% of the state's population and supports the development of several municipalities.
To properly evaluate the water availability in the Ivinhema basin and to promote socioeconomic and environmental development, different criteria for water concession in the Ivinhema river basin were analyzed in this study.

Material and Methods
This research was carried out through quantitative methods, the collection of quantitative or numerical data was performed through the use of measurements of quantities and obtained through metrology, numbers with their respective units. These methods generated a set or masses of data, which were analyzed using mathematical techniques Research, Society and Development, v. 9, n. 10, e1969108391, 2020(CC BY 4.0) | ISSN 2525-3409 | DOI: http://dx.doi.org/10.33448/rsd-v9i10.8391 (Pereira et al. 2018 The Ivinhema basin is located in Mato Grosso do Sul and occupies an area of approximately 46,500 km 2 . It is located between latitudes 20° 51' S and 23° 14' S, and longitudes 52° 21' W and 55° 57' O. The Pardo river basin is in its northern border, and the Amambai river basin in the southern. The Maracaju mountain range and the Republic of Paraguay are west of the Ivinhema basin, and the Paraná River is east of it (Arai et al. 2012).
The main river of the basin is the Ivinhema river, and its main tributaries are the Dourados river on the right bank and the Vacaria river on the left bank. Twenty-five municipalities are part of the basin. Fifteen of them are integrally within the drainage area, and 10 only partially.
According to the Köppen Geiger classification, the climate of the region is type Aw (Peel et al., 2007), with annual average precipitation and temperature of 1,425 mm and 23.6 °C, respectively.
To estimate the minimum reference flows (Q7,10 and Q95), data from nine fluviometric stations (Table 1) were gathered. These stations belong to the hydrometeorological network of the National Water Agency (ANA), and the data are available on its Hydrological Information System. To estimate Q7,10 and Q95, years from 1973 to 2007 with more than 95% of the data were analyzed. To estimate the annual and monthly series of Q7 for each station, log-normal Research, Society and Development, v. 9, n. 10, e1969108391, 2020 (CC BY 4.0) | ISSN 2525-3409 | DOI: http://dx.doi.org/10.33448/rsd-v9i10.8391 5 probability distribution functions, namely Pearson III, log-Pearson III, and Weibull, were analyzed with two and three parameters. The Kolmogorov-Smirnov adherence test was used to select a probability distribution that was adjustable for each historical series, with 20% probability. The adjustment of the parameters for each distribution was performed using the moments method. After selecting the probabilistic distribution with the best fit to the minimum flow data, the values of Q7,10 were obtained. To calculate the Q7,10 of seasonal periods, the events of Q7 were obtained by restricting the data set to the seasonal period in question. To compute these flows, the Computational System for Hydrological Analysis (SisCAH, 2009) program was used (Sousa et al., 2009).
Q95was obtained from the permanence curve of each fluviometric station, based on daily data.Q95considers the portion of time during which a given flow is equalized or exceeded within the analyzed period. Therefore, the data series were organized in descending order, and the frequency associated with each flow rate was determined according to Eq. 1: Where Nqi is the number of events greater than or equal to the flow of order i; and NT is the total number of data in the sample.
Through SisCAH, Q95was obtained in monthly and annual scales for each historical series, according to the methodology described. To determine the permanence curve on a monthly basis, the same procedure as the one for the annual estimation was used, but with the data set restricted to a monthly scale.
Based on the estimated minimum reference flows, them monthly and annual flows were compared. The relative difference between water availability for concession based on monthly and annual minimum reference flows were calculated according to Eq. 2.
Where Dr is the relative difference in water availability, %; Qseasonal is the estimated flow on a monthly basis,m 3 s -1 ; and Qannual is the estimated annual flow,m 3 s -1 .
For each fluviometric station considered in the study, the values of Q7,10 and Q95 on monthly and annual bases were compared, in addition to the criteria for: 50% of Q7,10, 50% of Q95, 70% of Q7,10,and 70% of Q95; in the different time scales analyzed. The monthly behavior of the estimates and the magnitude of the discharge according to the time scale were compared to the annual flows.
The relative difference between the criteria of maximum allowable discharge for annual and monthly bases was based on volumes and calculated using Eq. 3.
Where Drcriterion is the relative difference between the criteria, %; V70%Q95 is the maximum permissible water volume for concession considering the ANA criterion, m 3 year -1 ; and V50%Q7,10 is the maximum permissible water volume for granting, considering 50% of Q7,10, m 3 year -1 .
Based on the flows considered for the monthly and annual concession, the total volume for each concession (annual and monthly) was calculated. The difference between the maximum permissible volume upon change from annual to monthly concession was calculated according to Eq. 4.
Where Drb is the relative difference between the monthly and the annual criterion, %.

Results and Discussion
The monthly variations of Q7,10 and Q95 throughout the year, the comparison with annual values, and the projection of different criteria for the concession (50%Q7,10, and 70%Q95 monthly and annual basis) are presented in Figure 1.
On average, the annual Q95 was 35.0% higher than the annual Q7,10. Considering the different flow criteria for concession in the nine stations evaluated, 50% of monthly Q7,10yielded a 75.9% higher flow for concession compared to 50% of annual Q7,10. Moreover, Research, Society and Development, v. 9, n. 10, e1969108391, 2020 (CC BY 4.0) | ISSN 2525-3409 | DOI: http://dx.doi.org/10.33448/rsd-v9i10.8391 7 70% of monthly Q95resulted in a 32.0% higher result when compared to 70% of annual Q95. For all studied stations, it is evident that, compared to the monthly criteria, the criterion corresponding to the concession of 70% Q95 per year is very permissible in the months with low water availability (September to November) and quite restrictive in the months with high water availability (Figure 1). At the Retriat Guarujá station ( Figure 1E), 70% of annual Q95 and 70% of monthly Q95 present similar values for October and even higher results for May, compared to the monthly Q7,10. These results indicate that annual flows may not be a satisfactory criterion for water concession in the region, as it may negatively affect water availability. According to Bof et al. (2013), these circunstances can increase the risk of occurrence of unfavorable conditions that may lead to the complete drought of a river. If the 70% Q95 annual concession was adopted, the river would dry in October of that year at Aroeira station ( Figure 1F). However, if this same criterion is adopted on a monthly basis, the risk of river drought decreases.
The period from September to November is the most critical in terms of water availability. During these months, 70% of annual Q95 is close to the values of monthly Q7,10.
However, this risk of drought decreases if 70% of monthly Q95is used. At Farm São Joaquim station ( Figure 1G), 70% annual Q95resulted in a value above the monthly Q7,10 (10.3%)in September, which implies a high risk of river drought.
Overall, the criterion based on monthly flows leads to a more adequate water management plan, as it allows for a greater water use during the period of higher availability, and it imposes a more realistic restriction to critical periods. Bof et al. (2013) emphasizes the same argument in their paper about the Paracatu basin in Minas Gerais. Silva et al. (2011) state that seasonality in the granting process becomes increasingly essential in basins with high growth rates and potential conflict between users.
At the Aroeira and São Joaquim stations ( Figures 1F and 1G), 70% of annual Q95 is greater than the annual Q7,10. Therefore, the use of the former as the maximum permissible flow for concession would imply the complete drought of the river for seven consecutive days at least once every 10 years, sinceQ7,10 is the smallest and most restrictive reference flow. Bof et al. (2013) compared the maximum permissible flows for concession in the Paracatu basin, considering the criteria used by the Minas Gerais Water Management Institute (IGAM) and ANA on an annual and monthly basis. They observed that at the Farm Limoeiro station, the annual Q95 was 47% higher than the annual Q7,10. Therefore, 70% of Q95 would lead to a concession 3.4 times greater than the one based on 30% of Q7,10.
A concession based on 50% of annual Q7,10 would excessively limit the use of water resources throughout the year in the periods of high and low water availability. The change to 50% of monthly Q7,10would allow a higher amount of water to be used in periods with water surplus, and a lower amount in the months with lower water availability for all stations evaluated, as represented in Figure 1.
The permissible volume of water for concession is represented by the area under the curve (or line) relative to the criterion adopted. For the Ivinhema station ( Figure 1I), 50% of Q7,10 per year would allow the concession of a volume of 2,155.0 hm 3 . For 70% of annual Q95, the volume would be 3,863.2 hm 3 , 50% of monthly Q7,10would reach from 2,206.1 to 3,159.8 hm 3 .For 70% of monthly Q95, a maximum value of 4,591.6 hm 3 would be achieved.
These maximum values are 1.8, 1.5, and 2.1 times higher than the maximum volume allowed by 50% of annual Q7,10 (criterion assumed in several Brazilian states).
Drcriterion shows the relative percentage difference (%) of the total annual volume of water permissible for granting, between 50% of Q7,10 and 70% of Q95, on an annual and monthly basis, considering all nine fluviometric stations studied (Table 2). Research, Society and Development, v. 9, n. 10, e1969108391, 2020 (CC BY 4.0) | ISSN 2525-3409 | DOI: http://dx.doi.org/10.33448/rsd-v9i10.8391 Table 2. Percentage difference (%) of the total annual volume of water permissible for granting, between the criterion 50%Q7,10/70%Q95, on annual and monthly basis, in the 9 fluviometric stations analyzed. Considering the annual base, the volume of water granted based on 70% of Q95 is significantly higher than the one based on 50% of Q7,10, the relative percentage varies from 71.64 to 129.10%. On the monthly basis, the differences are lower and vary from 38.3 to 56.0%. These results show that the monthly Q7,10 and monthly Q95are somehow similar. Bof et al. (2013) obtained more expressive results for the Paracatu basin, since on anannual basis, the differences between permissible water volume adopted by ANA (70% of Q95) and IGAM (30% of Q7,10) ranged from 211.1 to 282.3%, and on the monthly basis this difference varied from 110.5 to 152.8%. Table 3 shows the percentage relative difference (Drb%) for the nine fluviometric stations analyzed. It presents the changes in the granting criteria upon a shift from annual to monthly basis. These changes were higher for 50% ofQ7,10, whose monthly flows increased up to 73.6%. In contrast, the shift from annual to monthly 70% of Q95 resulted in a variation up to 12.3% in the flow granted. In this context, the annual values are more restricted because they are based on the period of greatest water shortage of the year. Whereas the monthly analysis represents the intrinsic characteristics of the flows during each month. The Drb% was smaller for 70% Q95 because the flows of monthly Q95 were smaller than the annual Q95 in some months. Source: Authors. Bof et al. (2013) considered the change from the criterion by ANA (70% of annual Q95) to 70% of monthly Q95, which resulted in an increase in the maximum permissible volume for concession from 26.7 to 67.1%.
The graphs from Figure 2 show the amplitude of variation in the relative difference between monthly and annual Q7,10 and Q95flows, considering the fluviometric stations studied.
The monthly Q7,10 flows are higher than the annual Q7,10 in all analyzed months, characterizing a potential increase in the allowable flow for concession. This potential is accentuated from December to April, when the increase is higher than 40%, except in the Retriat Guarujá station, where Q7,10 monthly values were lower than annual Q7,10 in some months. For Q95 ( Figure 2B), negative values of relative difference (flow values in which annual Q95 is greater than monthly Q95) are observed from August to January and in May.
Although they imply a reduction in the maximum allowable discharge for these months, the use of Q95can increase environmental security, since the use of 70% annual Q95 leads to values approaching the monthly Q7,10.
The use of estimated minimum flows on an annual basis restrict the water use for the whole year. Moreover, the period of highest demand for water resources does not always coincide with the period of lower water availability (Bof, 2013). Therefore, there should be more studies on the quantification of water availability. In this scenario, a change on the concession criteria from an annual to a monthly basis can potentially increase the maximum allowable flow rate.

Conclusions
1. The replacement of the annual criteria (standard in Brazil) for water concession by 50% of monthly Q7,10 and 70% of monthly Q95can potentially increase the use and improve the management of water resources.
2. The best criterion for concession is the monthly Q7,10 because, despite being the most restrictive, it allows a higher withdraw in the periods with water surplus, and restricts the volume during the months with lower water availability.