*Appendix A.5. Multi-Source Weighted-Ensemble Precipitation Dataset (MSWEP)*

MSWEP is a newly raised global precipitation dataset (1979–present) with a 3-hourly temporal resolution. In order to acquire the best possible rainfall estimates, MSWEP is unique in its use of an unprecedented variety of data sources, such as gauges, atmospheric reanalysis models, and satellites. In brief, the main procedures carried out to generate the MSWEP dataset are composed of four steps. (1) The long-term mean of the MSWEP, which is mainly based on the Climate Hazards Group Precipitation Climatology (CHPclim) dataset, is bias corrected by using country-specific catch-ratio equations methods. (2) Several satellites (e.g., CMORPH, Global Satellite Mapping of Precipitation (GSMap MVK) and TRMM) and reanalysis (e.g., ERA-Interim, and the Japanese 55-year Reanalysis (JRA 55)) precipitation datasets are assessed in terms of their temporal variability to determine their potential contents in the MSWEP. (3) The long-term climatic mean is temporally downscaled in a gradual manner from monthly to daily by applying the weighted averages of precipitation anomalies from the gauge, reanalysis, and satellite datasets. (4) The long-term climatic mean is further temporally downscaled to the 3-hourly timescale to generate the final MSWEP dataset [71].

#### **Appendix B.** *Remote Sens.* **2020**, *12*, x FOR PEER REVIEW 28 of 34

**Appendix C** 

**Datasets** 

CN05

PERSIANN

and validation (2011–2013) periods.

**Calibrate by Dense-Gauge (NSE) 2004–2010** 

0.89

**Figure A1.** Winter precipitation (mm) during December–February of 2003–2013 from CN05 and eight satellite-based datasets compared with the dense-gauge dataset, which is shown as colored dots. **Figure A1.** Winter precipitation (mm) during December–February of 2003–2013 from CN05 and eight satellite-based datasets compared with the dense-gauge dataset, which is shown as colored dots.

**Calibrate by Satellite-Based Dataset (NSE)** 

0.86 0.89 0.88

**Validate by Satellite-Based Dataset (NSE)** 

**Validate by Dense-Gauge (NSE) 2011–2013** 

PERSIANN −0.4 0.3203 −0.154 CMORPH RAW −0.97 −0.66 −0.59 TRMM 0.73 0.79 0.72

CDR 0.56 0.6375 0.4783 CMORPH CRT 0.75 0.7815 0.7276 CMORPH BLD 0.84 0.8571 0.809 MSWEP 0.78 0.8517 0.799 CHIRPS 0.44 0.7019 0.42

**Table A1.** NSE of daily streamflow time series simulated by XAJ model calibrated by using eight

## **Appendix C.**

**References** 

*Hydrol.* **2016**, *542*, 343–356.

**Table A1.** NSE of daily streamflow time series simulated by XAJ model calibrated by using eight satellite-based datasets, CN05, and the referenced dense-gauged dataset for calibration (2004–2010) and validation (2011–2013) periods.


**Figure A2.** Mean monthly hydrographs (2004–2013) simulated by the XAJ model calibrated by using eight satellite-based datasets, CN05, and the referenced dense-gauged dataset. **Figure A2.** Mean monthly hydrographs (2004–2013) simulated by the XAJ model calibrated by using eight satellite-based datasets, CN05, and the referenced dense-gauged dataset.

2. Essou, G.R.; Arsenault, R.; Brissette, F.P. Comparison of climate datasets for lumped hydrological

3. Xu, H.; Xu, C.-Y.; Chen, S.; Chen, H. Similarity and difference of global reanalysis datasets (WFD and APHRODITE) in driving lumped and distributed hydrological models in a humid region of China. *J.* 

4. Essou, G.R.; Sabarly, F.; Lucas-Picher, P.; Brissette, F.; Poulin, A. Can precipitation and temperature from meteorological reanalyses be used for hydrological modeling? *J. Hydrometeorol.* **2016**, *17*, 1929–1950. 5. Lorenz, C.; Kunstmann, H. The hydrological cycle in three state-of-the-art reanalyses: Intercomparison

6. Beck, H.E.; Vergopolan, N.; Pan, M.; Levizzani, V.; Van Dijk, A.I.; Weedon, G.P.; Brocca, L.; Pappenberger, F.; Huffman, G.J.; Wood, E.F. Global-scale evaluation of 22 precipitation datasets using gauge observations

7. Blarzino, G.; Castanet, L.; Luini, L.; Capsoni, C.; Martellucci, A. 2009 3rd European Conference on Antennas and Propagation. In *Development of A New Global Rainfall Rate Model Based on ERA40, TRMM,* 

observations and numerical models. *Bull. Am. Meteorol. Soc.* **2007**, *88*, 47–64.

modeling over the continental United States. *J. Hydrol.* **2016**, *537*, 334–345.

and performance analysis. *J. Hydrometeorol.* **2012**, *13*, 1397–1420.

and hydrological modeling. *Hydrol. Earth Syst. Sci.* **2017**, *21*, 6201–6217.

*GPCC and GPCP Products*; IEEE: New York, NY, USA, 2009; pp. 671–675.
