General comments:

The manuscript assessed the effects from temporally explicit changes of climate variables and underlying surfaces on the streamflow trend using Variable Infiltration Capacity (VIC) model prescribed with continuously dynamic leaf area index (LAI) and land cover in Yellow River Basin. This study suggests that change in underlying surface has imposed a substantial trend on naturalized streamflow in Yellow River Basin. This topic is interesting and important for the water resources management in Yellow River Basin, esceically for soil and water conservation measures and ecological restoration projects.

Specific comments:

Line 156-159, the two-steps methed was designed to consider time variant LAI in the VIC model simulation. Is it possible to use interannual change of LAI and land use in VIC model? In many hydrolgical models, the dynamical LAI and land use data are used. The version of the VIC model should be introduced.

Line 189, The monthly stramflow is evlauated by NSE, Bias and RMSE, which should be stated here, and calibration period and validation period, too.

Line 281, the vegetation degradation in the source region and urbanization in the middle reaches. It’s better to cite reference or data to support this attribution.

Line 315, the grey frames in Figure 7 are not necessary.

Line 342, add Table 3.

Line 335-346, a table that summarizes the value of NSE, RMSE and Bias at different gauses in different period is helpful.

Line 338, maybe the calibtated values of 6 soil parameters mentioned in Line 185 should be presented here.

Line 353, “For the HYK station, the contributions of all climate variables to the streamflow trend were positive excepting temperature, while larger negative effects from underlying surface change offset the slight positive effects of climate change on the streamflow trend (Figure 9).”, it’s better to move this sentence to Line 361 after the simulation result.

Line 352-380, a table is needed to summarize the value of impacts and relative impacts rates shown in Figure 9.

Line 500, the slope land changes into the flat terraces could dramatically decrease the surface runoff generation and should not be ignored. It will also induce the change of intra-annual temporal pattern of LAI as shown in Figure 11(d).

Line 519, Was the degradation of permafrost simulated in VIC model in this study? How about the setting? Please explain it.

Based on the Variable Infiltration Capacity (VIC) model prescribed with continuously dynamic leaf area index (LAI) and land cover, this study attributed the trend of naturalized streamflow to temporally explicit vegetation change and climate variation over the Yellow River Basin of China. They found that the effect of climate variation on streamflow is slight, while the change of underlying surface has imposed a substantial trend on naturalized streamflow. This research is of significance for understanding the underlying mechanisms of natural streamflow reduction, which can provide guidelines for local water resources management. Here are a few comments below:

1.    Equation 7: In scenario S3 (f (Cinter, Pintra)), all climate variables and intra-annual temporal pattern of monthly precipitation vary according to observation records, while in scenario S2 (f (Cinter)), only specific climate variable varied according to observation records, why the intra-annual temporal pattern of precipitation on the annual streamflow trend (QPintra) can be calculated with equation 7? Maybe you should add a scenario (S2-1), in which all the interannual change of climate variables vary according to observation records while other variables vary according to control conditions in the S1. With this scenario, you can also check whether climate variables affect each other. 

2.    It seems that the VIC simulations don’t match well with the observations, and the Nash–Sutcliffe efficiency (NSE) of monthly streamflow is only 0.44 and 0.46 over TNH-TDG and LM-HYK during validation periods (Fig. 8). To ensure the accuracy of this research, it may be necessary to recalibrate the model parameters.

3.    Why only the results over TDG-LM and LM-HYK are shown in Fig. 12, and what’s the streamflow trend over other regions?

4.    To reduce the uncertainty of this research, it’s better to show the multi-years average evapotranspiration and soil moisture in Fig. 13, rather than a specific year. In addition, please explain the meaning of these line charts in the manuscript. 

Minor Comments

5.    Line 29: “the effect climate variation on streamflow” should be changed to “the effect of climate variation on streamflow”.

6.    Line 129: “hman” should be changed to “human”.

7.    Line 267: “HKY” should be changed to “HYK”.

8.    Fig. 4a: The label of the colorbar is incorrect. “-6” should be changed to “6”.

9.    Are the trends in Fig. 7, 10, and 12 significant? It’s better to add the confidence interval in the figures.

10.    Fig. 12: “LM-TDG” should be changed to “LM-HYK” in the figure caption. 

Attributing hydrological variation to climate and surface feature is a hot topic under global warming in Anthropocene and the Yellow River basin is a typical case where eco-hydrological processes have changed severely during recent decades, so it's critical to quantify the contributions of land use/cover and climate changes to streamflow reduction in the YRB, China. Many previous studies have already focused on this topic using various methods and have got some interesting findings. But as authors of this MS mentioned, those studies analyzed the contributions of inter-annual change of driving factors to hydrological processes, while effects of the intra-annual change of climate and vegetation have not been examined. To solve it, this MS improved the VIC model by coupling time-series land cover and LAI remote sensing data to capture the cumulative effect of dynamic vegetation on the hydrological cycle, and designed six scenario simulation experiments to separate impacts of intra-annual changes of climate and vegetation from those of inter-annual changes and the interactive effects. Since topic of this MS is meaningful and innovative, methods are convincing, and results & discussions are reasonable, I recommend it to be published with revisions. Some specific comments are as follows:

Line 187-188, the optimization algorithm named MOCOM-UA is inconsistent with SCE-UA in fig. 2.

Line 313, unit of runoff coefficient is wrong in Fig. 7a. 

Line 365, as Fig. 9 shows, temperature has negative impact on streamflow in the source regions. This is inconsistent with my understanding of this region that higher temperature contributes to an increase in runoff due to its role in promoting glacier melting, although the authors discussed the negative impact in Section 5.4 and attributed it to permafrost degradation. 

Line 390 It's confusing to use rainfall and precipitation concurrently because they have different meanings and precipitation includes rainfall, snow, hail, etc.

Line 435-439 There are some grammatical errors in these sentences such as vegetation phenology rather than phenological.

Line 455 Error in name of y-axis, annual instead of anuual. 

Line 455 Differences between trends of annual streamflow with continuous and noncontinuous LAI changes is little visually, especially for Figs. 12b, c, and d, so adding significant level of these trends may be more convincing. 

Line 458 Error in figure name; Figs. 12c and 12d are for LM-HYK instead of LM-TDG.

Line 486-491 This paragraph should move to Section 5.2 to build a direct connection with streamflow change, so that take Section 5.3 as an additional methodological analysis to highlight the role of vegetation dynamics in streamflow trend.

Line 544 Usage of due to is wrong, please check it throughout this MS. 

Line 547 Grammatical error, accounts instead of account.

Line 555-557 Usage of disentangle is wrong, rewritten it, please.