**1. Introduction**

Meteorological data such as air temperature and precipitation are important inputs to hydrological models. With a common-knowledge "Garbage in, garbage out" approach, meteorological data of good quality are prerequisites to achieve good simulation results using hydrological models and thus further to achieve reasonable decision support based on model outputs [1,2]. The traditional and common sources of meteorological data are ground measurements from gauge stations; such point-based measurements are considered as the most accurate data over the limited representative areas. The modelers need measurements from a dense network of gauge stations to adequately characterize the spatial and temporal variability of meteorological variables at the basin scale [3]. The in situ data collection and maintenances are usually time-consuming and labor-/resources-intensive, the gauge stations are unevenly distributed, and overall the number of stations is declining at the global scale [1]. As a result, modelers often encounter the challenge to obtain sufficient in situ measurements, as they expect. In developing countries and remote areas, gauge

**Citation:** Liu, J.; Zhang, Y.; Yang, L.; Li, Y. Hydrological Modeling in the Chaohu Lake Basin of China—Driven by Open-Access Gridded Meteorological and Remote Sensing Precipitation Products. *Water* **2022**, *14*, 1406. https://doi.org/10.3390/ w14091406

Academic Editor: Renato Morbidelli

Received: 26 March 2022 Accepted: 26 April 2022 Published: 28 April 2022

**Publisher's Note:** MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations.

**Copyright:** © 2022 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https:// creativecommons.org/licenses/by/ 4.0/).

stations are very sparse and in situ measurements are not even available over some regions. Many scientific communities have been stressing the need for more in situ measurements; one of the feasible ways to meet this need is to promote innovations and multidisciplinary cooperation in designing low-cost monitoring devices and in developing or combining monitoring techniques. In this regard, some concrete efforts are underway, for example, the working group Measurements and Observations in the XXI century (MOXXI) was established in 2013 with the specific aims of targeting innovation in all realms of hydrological measurements from ground-based to remote sensing [4,5].

In recent years, various freely available gridded meteorological datasets at different spatial and temporal resolutions over the global or quasi-global scales have been developed and released to the public [6–11]. For example, the Climate Forecast System Reanalysis (CFSR) dataset is such a global meteorological dataset covering the 39-year period from 1979 to 2017. The CFSR data were produced by a global, high resolution, coupled atmosphere– ocean–land surface–sea ice system [10]. The meteorological variables include precipitation, air temperature, wind speed, relative humidity, and solar radiation. The China Meteorological Assimilation Driving Dataset (CMADS) is a country-scale gridded meteorological dataset containing the same types of data as CFSR, which used more measurement data in China and integrated the Climate Prediction Center (CPC) morphing technique (CMORPH) satellite-based rainfall product [12]. Several agencies have preprocessed CFSR and CMADS products to generate the datasets in the desired input format of the widely used hydrological model, i.e., the Soil and Water Assessment Tool (SWAT) model [13]. This makes these data sets very convenient to use for the modelling community [14–17]. There are also many gridded precipitation datasets such as the TRMM (Tropical Rainfall Measuring Mission) multi-satellite precipitation analysis (TMPA) product [18] and CHIRPS (Climate Hazards Group InfraRed Precipitation with Station data) precipitation dataset [19]. In recent years, with the rapid development of machine learning and especially deep learning techniques [20], the accuracy of gridded meteorological and precipitation datasets is expected to be improved dramatically in the near future [21].

The increasing availability of gridded meteorological datasets has attracted attention to use them in driving hydrological models for streamflow simulation [22–25]. As the accuracy of these gridded meteorological datasets varies among regions [2,26–30], it is necessary to evaluate these datasets before their application in specific areas. In this regard, many evaluation studies have been conducted to assess the performance of open-access weather data in hydrological simulations by using the best available gauge data as a reference [31,32]. For CFSR, it was found to be able to drive the hydrological model to yield satisfactory streamflow simulation in Lake Tana Basin (the upper part of the Upper Blue Nile basin) [16], the Bahe River Basin of the Qinling Mountains, China [33], four small watersheds in the USA, and the Gumera watershed in Ethiopia [19]. However, unsatisfactory results of streamflow simulation using CFSR as forcing data were also reported in the upstream watersheds of Three Gorges Reservoir in China [34] and two watersheds in the USA [28]. For CMADS, most evaluation studies using it as forcing data showed satisfactory results, such as those conducted in the Yellow River Source Basin located in the Qinghai–Tibet Plateau [17], the Lijiang watershed in South China [14], and the Jing and Bortala River Basin in Northwest China [35]. It is well recognized that a certain product's performance would vary from region to region, and evaluation of a certain product in various environments is essential for understanding its global performance [7].

This study focuses on the Chaohu Lake basin in the middle–lower Yangtze Plain, China. This region has been facing serious water pollution problems due to non-point-source pollution caused by intense agricultural activities (e.g., pesticide and fertilizer use). [36]. Watershed simulation and scenario analysis are expected to provide valuable instructions for water quality control and water resources management. As water is an important medium for mass transport, adequate modeling of hydrological processes is a prerequisite to characterizing the nutrient migration processes at the watershed scale [37]. Reliable meteorological input data are the premise of the hydrological model setup. Considering

that measurements from meteorological and rainfall stations are usually hard to retrieve for many reasons (e.g., data-sharing policy), it is necessary to find out whether open-access gridded meteorological data can meet the requirements of hydrological modelling. In this study, a subbasin of the Chaohu Lake basin, where measurements from meteorological and rainfall stations are relatively complete, was selected to evaluate the performance of two mainstream, open-access, gridded meteorological datasets (i.e., CFSR and CMADS) and three popular, satellite-based, gridded precipitation datasets (i.e., TRMM, CMORPH, and CHIRPS) in driving SWAT in streamflow simulation in this region. mance of two mainstream, open-access, gridded meteorological datasets (i.e., CFSR and CMADS) and three popular, satellite-based, gridded precipitation datasets (i.e., TRMM, CMORPH, and CHIRPS) in driving SWAT in streamflow simulation in this region. **2. Study Area**  The Chaohu Lake, located in Anhui Province, China, is the fifth largest freshwater lake in China and it is of great importance in terms of water resources and agriculture [38].

meteorological input data are the premise of the hydrological model setup. Considering that measurements from meteorological and rainfall stations are usually hard to retrieve for many reasons (e.g., data-sharing policy), it is necessary to find out whether open-access gridded meteorological data can meet the requirements of hydrological modelling. In this study, a subbasin of the Chaohu Lake basin, where measurements from meteorological and rainfall stations are relatively complete, was selected to evaluate the perfor-

*Water* **2022**, *14*, x FOR PEER REVIEW 3 of 23

#### **2. Study Area** The Fengle river is a main tributary of the Chaohu Lake Basin (Figure 1). The drainage

The Chaohu Lake, located in Anhui Province, China, is the fifth largest freshwater lake in China and it is of great importance in terms of water resources and agriculture [38]. The Fengle river is a main tributary of the Chaohu Lake Basin (Figure 1). The drainage area of the Fengle watershed is 1500 km<sup>2</sup> with elevations ranging from 7 to 455 m above mean sea level, and the main stream length is about 50 km. The land use types include agricultural lands (about 45%), forests (39%), built-up lands (10%), and water areas (i.e., ponds and rivers, 6%). There are no large cities or industry factories in this river basin. Based on the available gauge precipitation data during 2008–2014, the mean annual precipitation is 1096 mm/year. The inter-annual distribution of precipitation is uneven, with the most precipitation occurring in spring and summer. Based on gauged data between 2008 and 2014, the average daily maximum and minimum air temperature are 21.1 ◦C and 12.3 ◦C, respectively, and the daily mean air temperature is 16.7 ◦C. area of the Fengle watershed is 1500 km2 with elevations ranging from 7 to 455 m above mean sea level, and the main stream length is about 50 km. The land use types include agricultural lands (about 45%), forests (39%), built-up lands (10%), and water areas (i.e., ponds and rivers, 6%). There are no large cities or industry factories in this river basin. Based on the available gauge precipitation data during 2008–2014, the mean annual precipitation is 1096 mm/year. The inter-annual distribution of precipitation is uneven, with the most precipitation occurring in spring and summer. Based on gauged data between 2008 and 2014, the average daily maximum and minimum air temperature are 21.1 °C and 12.3 °C, respectively, and the daily mean air temperature is 16.7 °C.

**Figure 1.** Locations of the Fengle river basin, gauge stations, and the center points of grid cells of the gridded meteorological and precipitation datasets. **Figure 1.** Locations of the Fengle river basin, gauge stations, and the center points of grid cells of the gridded meteorological and precipitation datasets.
