**Kebing Chen 1, Shenglian Guo 1,\*, Jun Wang 1, Pengcheng Qin 2, Shaokun He 1, Sirui Sun <sup>3</sup> and Matin Rahnamay Naeini 4,\***


Received: 1 November 2019; Accepted: 28 November 2019; Published: 1 December 2019

**Abstract:** Standard impoundment operation rules (SIOR) are pre-defined guidelines for refilling reservoirs before the end of the wet season. The advancement and availability of the seasonal flow forecasts provide the opportunity for reservoir operators to use flexible and early impoundment operation rules (EIOR). These flexible impoundment rules can significantly improve water conservation, particularly during dry years. In this study, we investigate the potential application of seasonal streamflow forecasts for employing EIOR in the upper Yangtze River basin. We first define thresholds to determine the streamflow condition in September, which is an important period for decision-making in the basin, and then select the most suitable impoundment operation rules accordingly. The thresholds are used in a simulation–optimization model to evaluate different scenarios for EIOR and SIOR by multiple objectives. We measure the skill of the GloFAS-Seasonal forecast, an operational global seasonal river flow forecasting system, to predict streamflow condition according to the selected thresholds. The results show that: (1) the 20th and 30th percentiles of the historical September flow are suitable thresholds for evaluating the possibility of employing EIOR; (2) compared to climatological forecasts, GloFAS-Seasonal forecasts are skillful for predicting the streamflow condition according to the selected 20th and 30th percentile thresholds; and (3) during dry years, EIOR could improve the fullness storage rate by 5.63% and the annual average hydropower generation by 4.02%, without increasing the risk of flooding. GloFAS-Seasonal forecasts and early reservoir impoundment have the potential to enhance hydropower generation and water utilization.

**Keywords:** Yangtze River; cascade reservoirs; impoundment operation; GloFAS-Seasonal; forecast evaluation

### **1. Introduction**

The rapid population and economic growth in recent decades, along with climate change and variability, impose more stress on water resources and cascade reservoir systems. Reservoirs, as one of the most important components of the hydrologic system, play a significant role as water supply by altering natural streamflow across space and time [1,2], while mitigating the effect of extreme events [3,4]. Conventionally, during wet season, reservoir operators release water preferentially for

flood control [5,6] while storing water before the end of wet season to meet the demand for hydropower generation, navigation, and water supply. For full replenishment of storage, reservoir operators and academic researchers have highlighted the importance of reservoir impoundment operation and impoundment rules in several studies [7–10].

Reservoir operation rules are often used to provide guidelines for reservoir operators to determine the amount of controlled discharge. Among these rules, the reservoir impoundment rules are designed to refill the reservoir and raise the storage water level. The New York City rule [11] is among the early guidelines for reservoir impoundment and provides the probability of spills rather than the amount of spill to minimize the water shortage [12]. Since the development of the New York City rule, different types of reservoir impoundment operation rules have been developed and employed for various reservoir systems [7,13–17]. In the Yangtze River, which is one of the largest rivers in the world by discharge volume with huge reservoir storage capacity, the impoundment operation is complex and challenging. Water managers and stakeholders employ predefined impoundment rules for reservoir systems. These fixed rules, so-called standard impoundment operation rules (SIOR), are derived based on the historical flow records [7,9,18]. The SIOR is designed to reduce the flood control risk during the impoundment period. However, these fixed rules are unable to fully replenish the storage of the reservoir during dry years, which could lead to water shortage. On the contrary, flexible early impoundment operation rules (EIOR) allow the reservoir operators to start the impoundment process earlier and avoid unnecessary spills. However, employing EIOR requires information about the streamflow forecast to alleviate the risk of flooding.

The recent advancements of the meteorological and hydrological forecast systems provide an unprecedented opportunity for employing flexible operation rules rather than fixed ones for reservoir systems [19–21]. Combining seasonal meteorological forecasts with hydrological models at continental-scale has provided several continental-scale seasonal hydro-meteorological forecasting systems [22–24], such as the European Flood Awareness System [25], the Australian Government Bureau of Meteorology Seasonal Streamflow Forecasts [26], and the National Hydrologic Ensemble Forecast Service, USA [27]. Several studies have demonstrated that a skillful streamflow forecast can enhance the efficiency of water allocation systems to manage the trade-off between hydropower, irrigation, municipal, and environmental services [28–31]. The potential for employing seasonal forecast in the Yangtze River basin has been investigated in several research studies, mostly through statistical techniques [32–34]. However, the potential application of the available seasonal forecasts for reservoir impoundment operation is not well understood in the Yangtze River basin. In this study, we evaluate the global seasonal river flow forecasting system (GloFAS-Seasonal) developed by the European Centre for Medium-Range Weather Forecasts (ECMWF) [35] for reservoir impoundment in the upper Yangtze River basin.

We follow two steps for our evaluation. First, we investigate different streamflow thresholds to evaluate the possibility of employing EIOR. These thresholds can be considered as an indicator for the dry condition which has adverse effects on reservoir impoundment operation. To find these thresholds, we analyze multiple impoundment operation scenarios for EIOR and SIOR using a simulation–optimization model. We test different percentiles of historical streamflow as thresholds to find the impoundment rule curves for these scenarios. These rule curves are derived using the non-dominated sorting genetic algorithm-II (NSGA-II) [36] through a multi-objective optimization process. We then analyze the objective function values to select the most suitable thresholds for employing EIOR or SIOR. Second, we apply these thresholds to evaluate the skill of the GloFAS-Seasonal streamflow forecast for selecting EIOR or SIOR. When the streamflow forecast is below the selected threshold, it means a dry condition. Hence, the EIOR provides a longer impoundment period during this condition and a more suitable impoundment approach for water conservation. We employ different scores to show the skill of the GloFAS-Seasonal streamflow forecast to detect the streamflow condition according to selected thresholds.

The rest of the paper is organized as follows: In Section 2, after introducing the study area, we build a cascade reservoir impoundment model for reservoir early refill operation. Section 3 reviews the GloFAS-Seasonal forecasting system and explains different measures to evaluate the skill of the forecast for streamflow conditions. We demonstrate and discuss the results for the streamflow thresholds and performance of the GloFAS-Seasonal forecast in Section 4. Finally, we draw the conclusion in Section 5.
