BIORETENTION AS A CONTROL TO URBAN DRAINAGE SYSTEM WITH AN ECOHYDROLOGICAL BASE: GIS AS A TOOL ON DECISION MAKING

The occupation and use of increasingly impermeable urban land have made it difficult to infiltrate water and, 57 consequently, increase the volume of runoff in different cities, which has required the development of 58 bioretention techniques in the field of hydrology. The aim of this article is to define and apply criteria for the 59 identification of areas for the construction of Bioretention systems for evaluations based on Geographic 60 Information System indicators, considering the aspects of quantity and quality in urban drainage. The developed 61 method allows to verify and compare changes in the surface of urban areas and their interference in the local 62 environment, the mapping of land use and occupation to simplify procedures to define and prioritize areas for the 63 construction of Bioretention systems, the use of resources from georeferenced bases to resolve eco-hydrological 64 issues. The study develops technical bases for the use of a georeferencing tool to analyze areas with speed and 65 consistency as a basis for decisions on the implementation of Bioretention systems.

systems can be assembled with any type of geotextile to allow infiltration or include an 138 impermeable coating to help capture rainwater and reuse it (FAWB 2009). The captured 139 rainwater is treated through a variety of physical, chemical, and biological processes, such as, 140 for example, mechanical filtration, sedimentation, adsorption, absorption, and microbial plant 141 (MULLANE et al. 2015). However, there are still some difficulties to be overcome for the 142 implementation of bioretention techniques. These difficulties involve finding variables and 143 adjusting the system to them. These variables are, for example, close to water bodies, built 144 areas, conservation areas, and soil type. Besides, the points of contribution and the climatic 145 characteristics of the region must be observed and studied thoroughly to meet the standards 146 required to increase the efficiency of the bioretention process (DOKULIL 2016). 147 In parallel, ecohydrology emerges as a conceptual and practical tool that helps to 148 understand the complex interaction of hydrological mechanisms (ZALEWSKI 2002), such as, 149 for example, the establishment of connections between ecological paths and processes, use of 150 ecological systems, hydrological structures, and types of land use and occupation. for basins with less than three km². As the application of bioretention catchments occurs even 219 in the micro drainage level, the recommendations of the manual and the rational method for 220 calculating the drained peak flow were used. First, it was calculated from the total runoff 221 volume from the determination of effective precipitation, based on Eq. 1 and 2. 222

Eq.2
Where: P is the total precipitation in the área; Pe is the effective precipitation; S (mm) is the ground retention 225 potential, and CN is the coefficient number.

227
The CN value represents the soil coverage conditions and varies from a very 228 permeable cover (lower limit value = 0) to a completely impermeable cover (upper limit value 229 = 100). The hydrological group of the soil and its use and occupation were used as input data. 230 Eq. 3 provided the value of CN os for cases where there was more than one group of land or 231 use and occupation. 232 Where is the i-th portion of the basin which has the ; is the coefficient of the i-th portion 235 of the basin.

237 238
The peak drained flow was then calculated using the ratio method. The determination 239 of the runoff coefficient (C) results from the ratio between the total volume of precipitation 240 and the total volume drained, that is, the ratio between precipitation and effective 241 precipitation, as shown in Eq. 4 (KAWATOKO 2012). This proportion represents the 242 percentage of runoff generation for the contribution interest area. From C referring to the 243 contribution area and the intensity of the rains, it was possible to obtain the peak flow drained, 244 as shown in Eq. 5.  Table 1 shows the data for controlling the mass balance time and the sample 280 collection period to determine the concentration at points 1 and 5. 281 Table 1 282 283 The time control applied to this balance was the same as the duration of the entry 284 occurrence. However, this period is much longer than the time covered by the sample 285 collection for the concentration measurements. The Event Mean Concentration (EMC) (Eq. 7) 286 was the basis for calculating the total input and output masses of events 2 and 3 to overcome 287 this problem. 288 Where the C t and Q t are the analyzed concentration parameters and the flow rate, at time t, respectively, and 291 the interval of each sample collation.

293
In theory, the event load is calculated by Eq.8 and can be simplified to Eq. 4.9: 294 Eq.8 295

Eq.9 296
Where Q m is the mean flow rate, C the concentration and t c the time control.

298
3 Results and Discussion 299 The presentation of the results addresses the use of geoprocessing and defines the tools 300 used, the criteria for quantifying land use, the calculation of surface runoff from the study 301 area, the Bioretention based on eco-hydrological principles and, finally, the water quality in 302 bioretention systems. The image processing allowed the identification and quantification of the types of 335 land use and occupation on campus 2 at USP São Carlos (Table 2). 336 337 Table 2 338 The data in Table 2 show that there was an increase in the built and paved areas and 339 the forest cover area over the years. It is possible to infer that there was an increase in the 340 waterproofed area within the campus, with the increase in paved and built areas, going from 341 3.78% to 14.44% of the total area. The increase in the built-up area reflected in the increase in 342 soil sealing and, consequently, in the increase in water runoff. Therefore, there is a more 343 significant amount of sediment transported at a higher speed to the receiving water body, and 344 this can increase its level of sedimentation. The data obtained from the construction of maps to show land use and occupation at 360 Campus 2 show that this urbanized space showed moderately accelerated growth, with an 361 increase of about 10% in impermeable areas and, consequently, reducing areas with 362 vegetation. This demonstrates the need for future planning. 363 Table 3 364 365 Table 4 366 367 Table 3 and Table 4 present data on the historical evolution of the legal reserve areas 368 (RL) and permanent preservation areas (APP). These data will be used to establish a 369 correlation with the emerging patterns of photointerpretation. In technical terms, there was no  The mapped scenarios denote a significant increase of the impermeable area through 417 the classes that indicate constructed areas and by the objects that indicate road expansion. 418 Under these conditions, after local consultations, it can be noticed an increase in volumes of 419 water running into the river studied. Additionally, the hydrological performance can also be 420 compromised over time, because, besides the running, the cycle as a whole is influenced by 421 the permeability coefficients. The dimensions of the Bioretention system followed the criteria of general efficiency 426 and joint performance and can be verified by comparing the qualitative and quantitative 427 efficiency. In this work, only qualitative results will be shown, guided by eco-hydrological 428 principles, as shown in Figure 5, which is illustrative and is not on the scale. The diagram 429 shows the contribution areas of the hydrographic basin studied concerning the Bioretention 430 technique, the zones of influence that emerge from the urban growth of the region, that the 431 Mineirinho River has dense riverside vegetation and small pockets of water along its route, 432 and that the contribution flow is basically from the campus drainage system and rainwater 433 harvesting. All of these components are evaluated in correlation with ecosystem interactions. 434 440 441 The proposed system allows this temporal assessment to be scaled in a modular way, 442 and without much effort to expand its size over time. These conditions are necessary because 443 the study area is recent and is in full expansion, that is, the more urbanized space, the greater 444 the flow and, therefore, the greater the Bioretention system.  Table 6 also shows, for 453 means, characteristics of events. 454 455 Table 6 456 457 The Table 7 shows the EMC obtained for the variables: Fe, Zn, Pb, Ni, Mn, Cu, Cr 458 and Cd, and shows, for means of comparison, the standards established for the effluent 459 released in a Class 2 river, which is Mineirinho's class, according to Brazilian Federal 460 CONAMA 357/2005 resolution. It can be observed that the outlet EMC for Fe, Ni, Cu, and 461 Cd, at all three events, and for Pb, in events 2 and 3, presents higher values than the limits 462 established by the CONAMA. However, the total inlet and outlet mass of the parameters 463 should be analyzed to verify the bioretention efficiency and how this system reduces the 464 impact on the receiving water body. Table 7 466 467 The rainfall measurements were 6.0 mm in the first event, 3.0 mm in the second 468 event, and 39.0 mm in the third event. Figure 7 shows the accumulated and instantaneous 469 water depth, generated by these rainfalls for each event. It can be observed the infiltration 470 depth in the Bioretention, being the difference of the inlet and outlet the water depth. In event 471 one (1), the infiltration rate was approximately 34%, to event two (2) 41% and, to event three 472  The inlet and outlet mass of each parameter analyzed, and the removal efficiency is 478 shown in Table 8. It shows the removal efficiency for the monitored events. It is possible to 479 notice, for most of the parameters analyzed, that the practice decreases the pollutant load, 480 which would have reached the water body, considerably. The chrome concentration measured 481 was insignificant, and it is not shown in the table. 482 The metal removal efficiency varies from 40.3% to 97.8%. The lead was the 483 parameter with the higher removal rates (94.0% and 91.8% for the events 1 and 2). Iron had 484 the worst removal efficiency: 40.3% in event 1, and it had even, for event 3, a higher outlet 485 mass than inlet, resulting in a negative efficiency (-74.84 %). This can be explained due to the 486 Brazilian soil chemical composition. Due to the geological characteristics and its high 487 weathering levels in tropical regions, it is common to find high iron oxides in the soils like the 488 one used to the superficial and vegetated layer of the CT in which an erosion occurrence was 489 observed during the event 3. Besides the iron, all other metals, in event 3, had a lower 490 removal efficiency when compared to the events 1 and 2. The reason might also be linked to 491 erosion. 492 Table 8 494 495 To The increase in impermeable areas, even with a slow rate of advancement, increases 533 the quantities of sediments during periods of heavy rainfall, making a high input of sediments 534 in the receiving body, causing the increase of various parameters related to water quality and 535 peak flow. With the implementation of compensatory techniques, this problem is mitigated, 536 lowering the impact in water ecosystems downstream of the bioretention technique. 537 The quality of the output effluent from the bioretention system confirms a significant 538 improvement in the analyzed parameters. It was possible to verify that there is a decrease in 539 the mass loading and, accordingly, improvement in water quality parameters, which 540 contributes to the lower of pollutants and contaminants to the receiver. This feature reinforces 541 the capacity of the treatment of the bioretention technique. 542 Guided by our studies, we can say that applying this methodology to select areas to 543 implement the bioretention system is efficient and proves, with the data presented, the reduction of pollutant load directed to the river, being understood that the construction of 545 bioretention techniques, in addition to not impacting the river, brings benefits to it. This is 546 justified by the thematic maps, where one can see that there is considerable natural vegetation 547 covering the area, as well as areas that were reforested over time. As for the quality of the 548 output effluent from the bioretention system, it is possible to verify that there is a significant 549 improvement of the analyzed parameters. 550 The study clearly shows the importance of using a georeferencing tool for analysis of 551 areas since it enables rapid and consistent analysis of the study area. When correlation occurs 552 with ecohydrological indicators, the perception of the process efficiency supported with 553 physical elements becomes believable when it is noticed. An example is the investigation of 554 vegetal areas, that together with the results of qualitative parameters, demonstrate that the 555 implementation of a bioretention system attenuates the peak flow effects for a specific region. 556 Studies of this size can support a feasibility analysis and is more accurate in efficiency for 557 regions with different climatic and physical characteristics. Competing interests 583 The authors have no competing interests to declare. 584 585 9 Availability of data and materials 586 The data used in this study are not organized in a parameterized database in a relational 587 bank for availability for automatic processing. However, to databases in multiple 588 spreadsheets, they can be made available, in .xlsx format, upon request by email: 589 altair.rosa@pucpr.br 590 591