A Hydrological Model of the Mundaú River Basin, Recalibration of a MGB-IPH rainfall-runoff model using updated land-use data

Abstract (Swedish)

Avrinningsområdet Mundaú har sitt ursprung i den Brasilianska delstaten Pernambuco och rinner i sydöstlig riktning mot Mundaúlagunen längs Atlantkusten tillhörande delstaten Alagoas. Området faller under nationell jurisdiktion, är relativt oreglerat och erfar både översvämningar och torrperioder. Modellering av avrinningsområdets hydrologi kan ge en fördjupad förståelse av dessa händelser och hur förändringar i området kan påverka dess hydrologi. Denna studie syftar till att bygga en MGB-IPH hydrologisk modell av avrinningsområdet som kan användas för att avgöra huruvida gällande tillstånd att använda vatten överstiger regleringar av tillåtna nivåer på månadsbasis baserat på ett genomsnittligt modellerat månatligt flöde. För detta syfte har MGB-IPH modellen kalibrerats för att närmast efterlikna uppmätta värden. Modellens simulerade flöden indikerar att de uppmätta värden som finns inte följer det förväntade mönstret för oreglerade flöden. Detta indikerar att det sker stora vattenuttag som påverkar uppmätta flöden nedströms. Dessa uttag kan ännu inte kvantifieras.

Abstract

The relatively unregulated, union domain Mundaú river basin, upper reaches originating in the Brazilian state of Pernambuco and flowing southeast to the Mundaú Lagoon along the Alagoas state’s Atlantic coast, suffers from both flooding and drought. Modeling the basin’s hydrology can lead to a deeper understanding of these events, and how changes within the basin can affect its hydrology. This study aims to build a MGB-IPH hydrological model of the river basin to be used to evaluate whether or not current water extraction permits for the Mundaú basin exceed the regulated flow (Q95) regulations – on a monthly basis – using a modeled monthly flow average. For this purpose, the MGB-IPH model has been calibrated to a best fit to observed data. The model’s simulated flows indicate that the observed data does not follow a pattern for unregulated flows in the basin. It is believed that there are large water extractions that affect the downstream data, extractions that cannot be quantified at the moment.

Popular Abstract

Computer modeling is often used to help describe nature when there is a lack of data. But even when the model’s results appear to be “good enough”, are they really reliable?

Many people around the world constantly live with dry periods with intermittent flash floods. The people living near the Mundaú river in northeastern Brazil are no different. Late night floods in 2010 for example reached heights of over 11 meters, decimating the sleeping towns and cities along the river’s banks. These types of catastrophic events have brought together researchers from Sweden and Brazil in order to try and better understand the local water balance in order to help insure the safety and wellbeing of the local people.

With having both a distinct wet and dry season, several different computer models have been developed and used in attempting to describe the natural water balance of the Mundaú river basin. Most models have concentrated on recreating flood events while this model has concentrated on river’s low flows. By examining the river’s low flows we are able to help in planning and administration of the basin’s water.

In this model it was found that even though the criteria for evaluating the model’s efficiency were acceptable, the model tended to over estimate the river water levels on the scale of 50-200%. Being calibrated to known observed flows in the river, many questions were asked about why such large discrepancies in flows were modeled. Is the model flawed in some way and if so are there any solutions to the flaws? How have other modelers that have used the same modeling program account for this flow issue? Are their flaws in the known data? Could known water extractions account for these differences in river flows?

The program used for modeling the Mundaú basin was developed to account for local Brazilian issues with, for example a lack of data, soil types and land uses. It has been used in modeling both large and small Brazilian river basins. Other modelers of the Mundaú basin have modeled the river’s high flows which are less susceptible to water extractions, thus they never encountered this issue.

An analysis of water extraction permits shows that permit allowances account for 25-50% of the water level differences. On the other hand, it was also found that the minimum level of water extraction needed in order to be required to attain usage permits is set at such a high flow for this river that there is a large possibility that many water users in this area are not required to request usage permits.

• author: Cockrell, Michael
• supervisor: Cintia Bertacchi Uvo; Carlos Ruberto Fragos
• organization: Division of Water Resources Engineering
• course: VVR820 20151
• year: 2016
• type: H2 - Master's Degree (Two Years)
• subject: Technology and Engineering
• keywords: Rainfall runoff modeling, MGB-IPH, Mundaú, Brazil, Water allocation grants, ANA
• publication/series: TVVR16/5002
• report number: 16/5002
• ISSN: 1101-9824
• language: English
• additional info: Examiner: Rolf Larsson
• id: 8858567
• date added to LUP: 2016-04-18 15:18:02
• date last changed: 2019-04-01 12:07:38