Satellite-based hydrological dynamics of the world's largest continuous wetland Natasha Costa Penatti a, ⁎, Teodoro Isnard Ribeiro de Almeida a , Laerte Guimarães Ferreira b , Arielle Elias Arantes b , Michael T. Coe c a University of São Paulo, Institute of Geosciences, Rua do Lago 562, Cidade Universitária, São Paulo, SP 05508-080, Brazil b Federal University of Goiás, Image Processing and GIS Lab (UFG/LAPIG), Goiânia 74001-970, Brazil c The Woods Hole Research Center, Falmouth, MA 02540-1644, USA abstract article info Article history: Received 5 August 2014 Received in revised form 24 August 2015 Accepted 31 August 2015 Available online xxxx Keywords: Pantanal wetlands Upper Paraguay River Basin Satellite-based water balance We investigate the potential for closing the water balance purely from remote sensing (RS) sources and quantify the hydrological dynamic of the Pantanal (Brazil), the world's largest continuous wetland. We use 10-year time series of total water storage changes (ΔS) derived from GRACE and the balance between precipitation (P) derived from TRMM and evapotranspiration (ET) derived from MOD16, as well as the overall vegetation response (EVI2) to water availability. The GRACE-estimates of total water storage were consistent with in situ measurements from the Ladário gauge station. Despite the coarse spatial resolution of GRACE, its estimates were able not only to represent the hydrological regime of the entire basin but also its internal variability. The total water storage change estimates correlated well with precipitation (r = 0.87), evapotranspiration (r = 0.83), and vegetation greenness (r = 0.85), particularly when a two to three month time lag was considered. Likewise, the MODIS- derived vegetation greenness was consistent with variations in precipitation (r = 0.77) and evapotranspiration (r = 0.79). Nevertheless, we found that the water balance could not be closed with these data. Inferred runoff was greatly overestimated due mainly to an underestimation of ET. The uncertainty in the inputs and scarce val- idation data were limiting factors. © 2015 Elsevier Inc. All rights reserved. 1. Introduction Wetlands play a significant role in the water cycle and are recog- nized as biodiversity hotspots (Mitsch & Gosselink, 2007) performing many vital ecological functions including: regulation of the hydrological cycle, flood control, improvement and maintenance of water quality, among others. The study of the general processes of the wetlands hy- drology is important for the development of sustainable wetland man- agement, emphasizing the maintenance of ecosystem services for the environment and society. Despite their importance, the hydrological dynamics of seasonally flooded wetlands and floodplains remains poor- ly quantified through ground observations, satellite observations or modeling (Lee et al., 2011). The quantification of the spatial and temporal changes of water bal- ance variables and water budget closure over large spatial scales is in- strumental to understand the availability of water resources. The terrestrial water budget consists of four main terms: precipitation (P), evapotranspiration (ET), runoff (Q) and total terrestrial water storage change (ΔS). Runoff can be measured by streamflow gauges and pro- vides data on watershed characteristics. However, it is a challenge to measure the other variables over large scales with on-ground observa- tions at reasonable costs due to difficulties in representing spatial het- erogeneity and sampling errors (Gao, Tang, Ferguson, Wood, & Lettenmaier, 2010). Satellite remote sensing (RS) can monitor over large spatial scales in near real time, providing observations of land surface hydrological fluxes, particularly in regions where in situ networks are sparse. Several recent studies have investigated the water budget closure from key hy- drological components acquired from space and compared multiple datasets in order to obtain more robust representation of the water fluxes. Most of them, conducted in basins with ample ground data, pointed out the difficulty to close this budget due to the lack of accuracy of the individual datasets and to their inconsistencies (Table 1). Data on water movements in the Pantanal, considered the largest contiguous wetland in the World (Alho, Lacher, & Gonçalves, 1988), are scarce. Because of data limitations, earlier hydrodynamic studies ei- ther focused only on small portions of the basin or used a simplified ap- proach (Bravo, Allasia, Paz, Collischonn, & Tucci, 2012). Thus, most of the studies of the Pantanal hydrology and hydrodynamics are based on scarce in situ gauge stations measurements (generally rainfall, river stage, and discharge data), particularly from the Ladário gauge, which has been maintained since 1900 by the Brazilian Navy to monitor the Paraguay River level at Ladário city. Hamilton, Sippel, and Melack Remote Sensing of Environment 170 (2015) 1–13 ⁎ Corresponding author. E-mail address: natasha.penatti@usp.br (N.C. Penatti). http://dx.doi.org/10.1016/j.rse.2015.08.031 0034-4257/© 2015 Elsevier Inc. All rights reserved. Contents lists available at ScienceDirect Remote Sensing of Environment journal homepage: www.elsevier.com/locate/rse