Land cover change explains the increasing discharge of the Paraná River
Over the past 40 years, the discharge in South America’s Paraná River basin has increased despite no evidence of significant rainfall increases in the basin. In this analysis, we show that the observed multi-decadal increase in discharge can be explained by concomitant changes in land cover that have occurred within the basin during this period. Our analysis also indicates that the peak discharge timing may have shifted concurrently from January/February in the 1970s to March in more recent decades. While land-use effect dominantly alters the long-term temporal dynamics of the river discharge over multi-decades, the change in the seasonality of the discharge can be attributable to the combined effect of the land-use and climate variability. This study suggests that the mean annual discharge is likely to change in the other South American River basins where land transformation is currently taking place, and the shift of the month of peak discharge needs to be taken into consideration to forecast the hydropower generation under changing climate and land conversion.
KeywordsLand cover change River discharge Paraná River basin Shift of the peak discharge timing Hydroelectricity generation in Brazil
Hosting multiple hydropower dams that generate a significant portion of electricity to meet the regional demand in southern Brazil, the Paraná River basin is geographically and economically important. The famous Itaipu hydropower plant, built in the lower reaches of the upper Paraná River basin at the Brazil-Paraguay border, has the installed capacity of 14,000 MW (MW), yielding the electricity that supplies 15% of Brazil’s total energy consumption (https://www.itaipu.gov.br/en).Considering the country’s heavy dependency on hydropower (up to 80%, U.S. Energy Information Administration 2014), the variability of the flow in the Paraná River basin is undoubtedly critical for the region’s sustainable energy supply.
Over the past several decades, the mean discharge of the Paraná River has increased notably (Tucci and Clarke 1998; Dai et al. 2009; Carvalho et al. 2011). Having the 1970s as the baseline period, the reconstructed natural flow inferred from the measurements from the gauge station at the Itaipu dam (25.43 S, 54.59 W) affirms the trend: + 11.1% in the 1980s, + 18.0% in the 1990s, and + 6.3% in the 2000s (data from ANA, the Brazilian National Water Agency, http://www2.ana.gov.br/). However, there is no evidence of any significant increase in rainfall over this period. The baseline precipitation of the 1970s, from the combination of reanalysis and observation-based datasets (Sheffield et al. 2006), indicates that mean annual rainfall decreased by 1.5% in the 1980s (or by 5.0% excluding the rainfall of the 1983 flood event as an outlier), increased by 4.2% in the 1990s, and then declined slightly by 1.0% in the 2000s.
On the other hand, the basin has undergone historically extensive land transformation. In the state of Paraná, forest cover decreased from 23.9% in 1965 to 5.2% in 1990, being replaced by annual crops since the 1970s (Tucci and Clarke 1998). The relationship between land cover and river discharge has been reported previously, including the historical debates (review in Andréassian 2004) and the early work by Bosch and Hewlett (1982). In particular, the alteration of river flow associated with deforestation in tropical basins was observed for a large number of watersheds (review in Farley et al. 2005; Oudin et al. 2008). For instance, Coe et al. (2009) suggested that the degree of vegetation removal and the deforestation rate of particular watersheds affect the discharge of the Amazon River basin. The land cover change was also shown to alter the discharge flows in its tributaries such as the Toscantins River (Costa et al. 2003), the Ji-Paraná River (Rodriguez et al. 2010; Rodriguez and Tomasella 2016), and the Xingu River (Dias et al. 2015; Panday et al. 2015). Despite the large body of literature that suggested the altered river flows of the individual drainage basins in South America, the historical change in discharge of the Paraná River has received little attention.
In this paper, we examined the mechanistic linkages between climate variability, land-use, and resulting river discharge in the Paraná River basin using a terrestrial biosphere model, the Ecosystem Demography version 2 (ED2) (Moorcroft et al. 2001; Albani et al. 2006; Medvigy et al. 2009). The model results were evaluated against the natural flows at Itaipu and the regional total water storage (TWS) change from the NASA’s Gravity Recovery and Climate Experiment (GRACE) satellite observation. We then demonstrated that land transformation indeed accounts for the decadal increases in discharge of the upper Paraná River basin that have occurred in the past 40 years, despite little or no change in the basin’s rainfall. The potential shift of the seasonality of the river discharge, which is another important indicator to measure the hydrologic change (Döll and Schmied 2012), is also explored.
The ED2 model calculates the water, carbon, and energy dynamics of the land surface. One of ED2’s distinguishing features is its ability to describe, in a physically consistent manner, the coupled water, carbon, and energy dynamics of heterogeneous landscapes. This ability to incorporate the land-use change and resulting heterogeneous patterns of land cover, and its impacts on the hydrological processes, ecology, and energy balance means that the model is ideally suited for investigating the combined impacts of rainfall trends and land-use change within the upper Paraná River basin.
The heterogeneous landscape of the region was represented by the mixtures of primary forest, secondary forest, and agricultural area. Four plant functional types (PFTs) were used: (1) early successional trees (fast growing, low density, water demanding), (2) mid-successional trees, (3) late successional trees (slow growing, high density, shade tolerant), and (4) C4 type grass. While a grid-cell shared a homogeneous meteorological forcing, evaporation fluxes were computed from the sub-divided, multi-layer canopy structure by size and age within a grid-cell. The stomatal conductance that regulates the plant’s transpiration was calculated by the water vapor deficit functions (Leuning 1995). A multi-layer soil model (Walko et al. 2000) was used to simulate surface runoff, soil moisture, and drainage. With the land-use disturbance rate updated annually, the model simulated the hydrologic changes associated with sub-grid heterogeneity of vegetation dynamics in the model time steps that span from seconds to 5 min (e.g., energy and water cycles) up to 15 min (e.g., photosynthesis and soil respiration). For further description of the ED2 model, please see Medvigy et al. (2009), and also Knox et al. (2015) and Swann et al. (2015) that used the ED2 model’s couple version to a regional climate model to investigate the past, current, and future land-use impact on hydroclimate of the surrounding regions in South America.
Study domain and experimental design
Five simulations conducted in this study
Initial land cover
Time series of annual land-use transitions applied for 1970–2007
1970–2008 (Sheffield et al. 2006)
Land cover state corresponding to the land-use pattern of the year 1970
1970–2008 (Sheffield et al. 2006)
Land cover state corresponding to the land-use pattern of the year 2008
1970–2008 (Sheffield et al. 2006)
Time series of annual land-use transitions applied for 1970–2007
Cyclically repeated the 1970s climate
Time series of annual land-use transitions applied for 1970–2007
Cyclically repeated the 2000s climate
In the initial simulation (LCC-fullClim), the model was forced with both the observed 1970–2008 time series of meteorological forcing specified from the dataset by Sheffield et al. (2006) and the abovementioned time series of land-use transitions that occurred during this period. The contributions of changes in climate forcing and land-use change to the pattern and magnitude of discharge at Itaipu were estimated by conducting two additional pairs of simulations. In the first pair (1970LC-fullClim and 2008LC-fullClim), the model was forced with the observed time series of meteorological forcing for the period of 1970 to 2008 specified from the dataset by Sheffield et al. (2006) but with the land cover states corresponding to either the land-use pattern of the year 1970 (1970LC) or present-day land-use patterns (2008LC). In the second pair of simulations (LCC-70sClim and LCC-00sClim), the model was forced with the Hurtt et al. (2006) time series of land-use transitions that occurred during the period 1970–2007, with either climate specified for the 1970s (LCC-70sClim) or for the 2000s (LCC-00sClim). The 2008 climatology was applied for the 10th year of each decade. In LCC-70sClim and LCC-00sClim simulations, land conversion rates were applied to introduce disturbances originating in human activity specified from the Global Land Use dataset, for 1970–1999 from Hurtt et al. (2006). We applied the rate of year 1999 onward until 2008 because of little change in total agricultural area for 2000–2008 within the basin. The LCC-70sClim was forced by the climatological condition of the 1970s, and the LCC-00sClim by the climatology of the 2000s. All simulations generated 39-year multi-decadal monthly time series on the predicted changes in the water cycle and its major components (i.e., patterns of water discharge, storage, and evapotranspiration) via the coupled water, carbon, and energy dynamics across the basin.
Datasets and model evaluation metrics
The spatially integrated monthly discharge fluxes predicted by this scenario were then evaluated against the estimated natural flow at the Itaipu dam gauge station (25.43 S and 54.59 W). The natural flow was reconstructed from the daily discharge measurements at the Itaipu Dam, by taking into account the temporal gains and losses arising from reservoir operation and water withdrawals at the upstream of the gauge station (data available from the ANA (Brazilian National Water Agency) website (http://www2.ana.gov.br/). In addition, the model’s predictions of the temporal dynamics of total water storage (TWS) within the upper Paraná River basin were evaluated against the GRACE satellite measurements of this quantity (Rodell et al. 2004; Güntner 2008; Syed et al. 2008; Han et al. 2009). Because GRACE measures only relative differences in TWS, the respective mean values of TWS from model simulations were removed, and the anomalies were compared against the change in GRACE-estimated TWS over the period of 2002–2008. For large river basins such as the Amazon, the speed of the river flow must be factored into TWS storage calculations because the amount of laterally transported water within the basin is significant (Han et al. 2010). However, the relatively small size of the upper Paraná River basin and the monthly time scale of the analysis meant that river routing effects can be neglected, and thus the model’s predictions of TWS can be simply calculated based on the differences in the aggregate fluxes of water (i.e., precipitation minus evapotranspiration and runoff) within the river basin.
Model mean annual discharge validation
The corresponding predictions of Total Water Storage (TWS) within the basin are shown in Fig. 4b. As can be seen in the figure, the model’s predictions (purple line) agree with GRACE observations of TWS anomaly (black dots) in timing and amplitude of seasonal peaks and troughs from 2002 to 2008. The Nash-Sutcliff goodness-of-fit value, a measure of degree of model’s agreement with observation, between the LCC-fullClim and GRACE, is 0.6, indicating plausible matches between the ED modeled flow and the GRACE observation. The model reproduces well the observed inter-annual pattern of the seasonal change in TWS, such as larger peak-trough changes in 2003 and 2007 compared to those in 2004–2006, and the prolonged seasonal increase in storage that occurred in 2004.
Land-use impact on mean annual discharge increase over the last decades
The natural flow at Itaipu (green dot in Fig. 1) showed that in the 1970s baseline period, the annual discharge was 10,474 m3/s, but increased to 12,677 m3/s in the 1980s (a + 21.1% increase, or excluding the anomalously large 1983 flood year, a + 11.1% increase), to 12,138 m3/s the 1990s, and to 11,135 m3/s in the 2000s—increases of 18.0 and a 6.3%, respectively (Table A1). For the 39-year mean annual discharge over the period of 1970–2008, the increasing trend is not statistically significant because of the leveled-off discharge after 1998; however, if we consider only the non-PDO period (1970–1998), the Mann-Kendall test shows that the discharge increase remains statistically significant with a confidence level of 95%. According to Dai et al. (2009), this suppression may be the consequence of a large-scale climatic phenomenon, the Pacific Decadal Oscillation (PDO). If we consider all the available years (1931–2010) of the discharge measurement at Itaipu, the increasing trend of the annual discharge is statistically significant with a confidence level of 95%. This reaffirms the results by Dai et al. (2009) and Carvalho et al. (2011); nevertheless, the domains of their studies are slightly different from this work.
The predicted rates of discharge after the mid-eighties are considerably smaller in the absence of land cover change, 73.9% of discharge assuming no land cover change (9047 m3/s from the 1970LC-fullClim and 12,237 m3/s from the LCC-fullClim) in the 1990s, and 64.5% discharge assuming no land cover change (6210 m3/s from the 1970LC-fullClim and 9627 m3/s from the LCC-fullClim) in the 2000s, which suggests that land cover change has contributed to an increased rate of discharge of the basin and that climate variability alone cannot explain the observed increases in rates of discharge that have occurred since the mid-eighties (Fig. 5). On the other hand, assuming the absence of the land-use disturbance, the climate variability over the historical period causes discharge to slightly decrease in the 1980s (− 1.0%), increase in the 1990s (+ 27.9%), and decrease again in the 2000s (− 12.2%) compared to the rates of discharge during the 1970s (Table A2).
Seasonality shift of mean monthly discharge
Our results indicate that the mean annual discharge increase recorded over the past few decades in the upper Paraná River basin is primarily attributable to the expansion of agricultural land as a result of conversion from forests (Fig. 5). The amount of annual discharge increases about 28% (1970LC-fullClim vs. LCC-fullClim) with the 4.4 times expansion of the agricultural land (from 17 to 79% in Fig. 2a). Our findings agree with previous studies showing that land conversion from forest to agriculture can increase the discharge: at a small watershed in Upper Xingu River basin, land conversion from forest to soybean cultivation increased mean daily discharge about four times, as well as the variability between the wet season and the dry season (Hayhoe et al. 2011), and in a comparison study of two catchments in eastern Amazonia, the pasture catchment had higher discharge than the forest catchment (de Moraes et al. 2006). Our findings are also consistent with the studies using dynamic global vegetation models, which showed that the land-use effects have had increased runoff of the twentieth century in eastern Brazil (Gerten et al. 2008) and the tropics (Piao et al. 2007).
The shift of the month of the maximum discharge (Fig. 6) corresponds to the positive increase in rainfall in earlier months (February and March, see Fig. A3). The intensification of the wet season precipitation and the peak river discharge over the past several decades was shown over the Amazon River basin (Gloor et al. 2015), as well as the altered rainfall seasonality, for example, the delayed onset of the wet season at Rondonia in eastern Amazon (Butt et al. 2011) and an increase in dry-season length over southern Amazonia (Fu et al. 2013). Our results suggest that the inclusion of both land conversion and the variability and trends in climate explain the observed increase in the monthly and seasonal discharge anomalies in the Parana River basin (Fig.6 and Fig. 7).
Associated with the land clearing from 1970s to 2000s in our simulations (Fig. 2), the basin-wide mean monthly ET is predicted to decrease by the range of 0.1 to 0.6 mm per day (Fig. A4). In the ED2 model, the root depth decreases from 2~3 m for trees to 1 m for agricultural land, thus limiting the available water supply from deep soil, regulating the size of open stomata, and reducing the transpiration flux (a component of ET). The reduction in ET, triggered by the changes in root depth, may lead to a greater amount of discharge. There are other changes that occur with land transformation, such as changes in Leaf Area Index (LAI), biomass, and PFTs that can also affect ET. Note that the effects of irrigation are not included in the ED2 simulations. Because the regional agriculture of the Paraná basin is mostly rainfed and irrigation is applied to the very limited part of the basin, our model discharge is directly comparable to the natural flow. However, for the regions where the irrigation plays a major role in growing crops, the plants’ water stress is greatly reduced and the change in ET with land conversion is not expected as much as our result is demonstrated. For such regions, the relationship between the ET change and the discharge variation, associated with the land cover change, have to be more carefully interpreted by considering the impacts of irrigation.
The reduced ET associated with land conversion may also contribute to the precipitation reductions that have occurred in the neighboring regions of southern and southeastern Amazonia (Doughty et al. 2012; Gloor et al. 2013). Studies using the coupled version of the ED2 model that represents the water and energy feedback processes (ED-BRAMS model) suggest, however, that historical land conversion of the northern South America is not necessarily related to the regional pattern of the precipitation anomaly (Knox et al. 2015). A recent coupled biosphere-atmosphere modeling study indicates that the feedback arising from future land cover change over Amazonia does not immediately trigger the precipitation change, but rather alters the status of the atmosphere more suitable for convection (Swann et al. 2015). Therefore, the ET reduction of the upper Paraná River may contribute to the continental-scale feedback to the atmosphere; however, the extent and magnitude of the precipitation feedback arising from the land transformation analyzed in this study is subject for future research.
From an applied perspective, these changes in discharge of this basin have important implications for the hydropower generation potential of the Paraná River basin. For example, assuming a business-as-usual deforestation scenario, the effect of increasing mean discharge on the enhanced hydroelectric energy generation was predicted to be about 10% for a hydroelectric dam on the Xingu River (Stickler et al. 2013). On the other hand, however, the altered seasonality of the discharge also affects hydropower generation and offsets the advantage of the increased mean annual discharge because in the wet season, once the water level reaches the maximum dam storage capacity, the excess water must be released to prevent from spilling over and cannot be used for electricity generation. If a delayed onset and shortening of the wet season (e.g., Fu et al. 2013) occurs, the number of days for stable operation of hydroelectricity generation reduces and a sustainable power generation declines.
The ED2 model of this study is shown to reproduce reasonably well the observed discharge increase at Itaipu and the regional TWS derived by GRACE satellite measurement. This study affirms the significance of land-use change effects for estimation of the regional water budget of the South American River basins, and therefore contributes to the scientific community’s ongoing efforts to understand the impacts of land-use and climate change on the future variability of the South American hydrology. The terrestrial feedback of the water and energy cycles onto the atmosphere can alter the status of the atmosphere and potentially the rate and pattern of rainfall: both of which, in turn, can indirectly influence the aggregated values of regional discharge and water storage. Our result on the altered mean annual volume and the seasonality of discharge in the Paraná River basin, which has a history of an intensive land conversion over the past decades, suggests reevaluation of the stationary assumption on rainfall and discharge seasonality that has been applied in planning the hydropower plants in other river basins in South America. In the era of unprecedented large-scale land conversion and climate change, the complicated, and often paradoxical, response of the hydrologic cycle can be unraveled with the help of the terrestrial biosphere models that incorporate both land cover and climate change processes.
The authors thank for the generous financial support of the Sustainability Science Program (SSP) at the Kennedy school of government (HKS) at Harvard University. Shin-Chan Han was partially supported by the NASA GRACE project. The initial input land cover maps were processed by Marcos Longo and Abby Swann. The domain map was prepared by Stacy Bogan in the Center for Geographic Analysis (CGA) at Harvard University. The output files of five ED2 simulations are available from the authors upon request. The authors also thank Fabio Farinosi, Fabio Pereira, Mauricio Arias, and In-Young Yeo for helpful discussion and comments, and Erin Ciccone for proofreading.
- Camilloni I, Barros V (2000) The Paraná River response to El Niño 1982–83 and 1997–98 events. J Hydrometeorol 1:412–430. https://doi.org/10.1175/1525-7541(2000)001<0412:TPRRTE>2.0.CO;2
- Fu R, Yin L, Li W, Arias PA, Dickinson RE, Huang L, Chakraborty S, Fernandes K, Liebmann B, Fisher R, Myneni RB (2013) Increased dry-season length over southern Amazonia in recent decades and its implication for future climate projection. Proc Natl Acad Sci U S A 110:18110–18115. https://doi.org/10.1073/pnas.1302584110 CrossRefGoogle Scholar
- Gloor M, Barichivich J, Ziv G, Brienen R, Schöngart J, Peylin P, Ladvocat Cintra BB, Feldpausch T, Phillips O, Baker J (2015) Recent Amazon climate as background for possible ongoing and future changes of Amazon humid forests. Glob Biogeochem Cycles 29:1384–1399. https://doi.org/10.1002/2014GB005080 CrossRefGoogle Scholar
- Han S-C, Yeo I-Y, Alsdorf D, Bates P, Boy J-P, Kim H, Oki T, Rodell M (2010) Movement of Amazon surface water from time-variable satellite gravity measurements and implications for water cycle parameters in land surface models. Geochem Geophys Geosyst 11:Q09007. https://doi.org/10.1029/2010GC003214 CrossRefGoogle Scholar
- Hayhoe SJ, Neill C, Porder S, Mchorney R, Lefebvre P, Coe MT, Elsenbeer H, Krusche AV (2011) Conversion to soy on the Amazonian agricultural frontier increases streamflow without affecting stormflow dynamics. Glob Chang Biol 17:1821–1833. https://doi.org/10.1111/j.1365-2486.2011.02392.x CrossRefGoogle Scholar
- Hurtt GC, Frolking S, Fearon MG, Moore B, Shevliakova E, Malyshev S, Pacala SW, Houghton RA (2006) The underpinnings of land-use history: three centuries of global gridded land-use transitions, wood-harvest activity, and resulting secondary lands. Glob Chang Biol 12:1208–1229. https://doi.org/10.1111/j.1365-2486.2006.01150.x CrossRefGoogle Scholar
- Leuning R (1995) A critical appraisal of a combined stomatal-photosynthesis model for C3 plants. Plant Cell Environ 18:339–355. https://doi.org/10.1111/j.1365-3040.1995.tb00370.x CrossRefGoogle Scholar
- Moorcroft PR, Hurtt GC, Pacala SW (2001) A method for scaling vegetation dynamics: the ecosystem demography model (ed). Ecol Monogr 71:557–586. https://doi.org/10.1890/0012-9615(2001)071[0557:AMFSVD]2.0.CO;2Google Scholar
- Rodell M, Famiglietti JS, Chen J, Seneviratne SI, Viterbo P, Holl S, Wilson CR (2004) Basin scale estimates of evapotranspiration using GRACE and other observations. Geophys Res Lett 31:L20504. https://doi.org/10.1029/2004GL020873
- Stickler CM, Coe MT, Costa MH, Nepstad DC, McGrath DG, Dias LCP, Rodrigues HO, Soares-Filho BS (2013) Dependence of hydropower energy generation on forests in the Amazon Basin at local and regional scales. Proc Natl Acad Sci U S A 110:9601–9606. https://doi.org/10.1073/pnas.1215331110 CrossRefGoogle Scholar
- Walko RL, Band LE, Baron J, Kittel TGF, Lammers R, Lee TJ, Ojima D, Pielke RA, Taylor C, Tague C, Tremback CJ, Vidale PL (2000) Coupled atmosphere–biophysics–hydrology models for environmental modeling. J Appl Meteor 39:931–944. https://doi.org/10.1175/1520-0450(2000)039<0931:CABHMF>2.0.CO;2
Open Access This article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made.