Reply on RC2

Following your suggestion, we have revised the manuscript very carefully. The lists below are the responses to each comment and all the revising have been marked in RED in the marked-up mode (lines 17-20, 27-30, 34-38, 42-46, 53-57, 64-69, 75-92, 94, 113-115, 123-124, 126-128, 135-136, 139-141, 146-156, 169-175, 187-196, 200-201, 206-208, 215-217, 220-222, 228-232, 236-239, 246, 248-249, 255-256, 260-263, 268, 271-275, 277-279, 285-286, 295-297, 301-303, 306-311, 319-321, 337-338, 343, 356-358, 382-384, 387-389, 395-401, 416-417, 422-428, 436-438, 442-448). It is expected that the quality of the revised manuscript has been improved significantly and meet the demand of journal.

The main aims of the study is to develop a new approach, which could distinguish turbidity maximum zone (TMZ) in different estuaries and different seasons, and provide a new reference and fresh perspective for the study of TMZ.

Major comments
As the authors alluded to……

Response:
The studies of estuarine TMZ has a long history of over 80 years since it was proposed in the last century (Glangeaud 1938). The current definition of TMZ has been widely accepted and recognized across the world (Page 3, lines 53-57). As far as we are concerned, it is a much very challenging task for us to give a more rigorous and quantified definition for TMZ.
As you point out, measuring the accuracy of corresponding results is absolutely critical issues for the assessment of any methods. Following your thoughtful suggestion, the common accuracy measures of object extraction from remote sensing imageries, areabased accuracy measures (Cai et al. 2018), has been added to evaluate and compare the performance of the different methods (Section 2.4;Page 11,Page 16,Page 19,Page 24,Page 26,. Suppose that A E is the area of the extracted TMZ, A C is the correct part of A E , and A R is the reference TMZ (ground truth). Then the quality (Q) of the TMZ extraction results in the study could be defined as follow.
The range of Q is 0 to 1. The bigger Q the value, the higher the accuracy of TMZ extraction results and the better performance of the method.
In the PRE, all the TMZ extraction results based on TMZI and previous approaches has been assessed by the visual interpretation TMZ results ('ground truth') and the new added accuracy assessment measures. In order to illustrate the evaluation and comparison process better, some figures (Figs. 7, 8c, 8d and 11) in the manuscript are restructured to a sketch map here.
In the third row of the sketch map, regions indicated by yellow dashed frames are the visual interpretation TMZ results and defined as 'ground truth'.
In the fourth row, regions indicated by mango colors are the TMZ extraction results based on TMZI method. Fifth row shows the same results, but with the true color imagery as base map.
In the sixth and seventh rows, regions indicated by cyan colors and yellow colors are the TMZ extraction results based on Shi et al. (2017) and Wai et al. (2004), respectively.
On the one hand, the TMZ extraction results by TMZI and previous approaches have big difference, which could provide an intuitive and rough comparison. On the other hand, the quantitative accuracy assessment also showed TMZI model has a better performance. The main aims of the studies in the Hangjiang River estuary (HRE) and Moyangjiang River estuary (MRE) are to further assess the applicability of TMZI. The observation is also one of the most basic methods in geography and remote sensing research fields. Given few specialized studies in the two estuaries, the TMZ extraction results in the HRE and MRE were assessed mainly based on remote sensing imageries and the corresponding retrieved results.
We have revised the figures, results validation part, relevant contents and statements in the manuscript carefully (Page 11,.

Response:
It is right that the more validation and assessment, the better a new method. In this study, although the PRE, HRE and MRE are all in Guangdong Province, the three rivers and estuaries have significantly difference and its own characteristics (Pages 6-8, lines 123-124, lines 135-141).
The PRE is a horn-shaped estuary. The source of Pearl River is in Yunnan-Kweichow Plateau, southwest China. Pearl River is the fourth longest (2320 km) in China with a drainage area of 4.53•10 5 km 2 , and its annual runoff (3.26•10 11 m 3 ) is only smaller than Yangtze River. The sediment load of Pearl River is 7.53•10 7 ton/year. The HRE is a forking-shaped estuary. The source of Hanjiang River is in Zijin County, east of Guangdong. Hanjiang River has a length of 470 km and has the second largest drainage area (3.01•10 4 km 2 ) in Guangdong Province. The annual mean surface runoff of Hanjiang River is 2.45•10 10 m 3 with sediment load is 6.93•10 6 ton/year. The MRE is a calabash-shaped estuary. The source of Moyangjiang River is in Cloud Mountains, west of Guangdong. Moyangjiang River has a length of 199 km and a drainage area of more than 6•10 3 km 2 . The annual mean surface runoff of Moyangjiang River is 8.21•10 9 m 3 and sediment load is 3.27•10 5 ton/year.
Based on the characteristics of the three rivers and estuaries, it could be found that the source, length, drainage area, shape, runoff and sediment discharge of them are different from each other, Which has a certain representativeness in the study of estuary.
On the other hand, considering that the study of TMZ has long been a hot topic in many fields, such as estuary and coast, hydrology, ocean, ecology, environment, geography and remote sensing, it would be better and more persuasiveness if the model could be further validated and assessed by other researchers worldwide.

Response:
It is no doubt that there exists a connection between total suspended solids (TSS) and chlorophyll a (Chla), even other water parameters or color components (CDOM). We have modified the relevant expressions to make it more precise (Page 2, lines 27-30; Page 5, lines 94-102; Pages 13-14, lines 220-222; lines 236-239).
In fact, the defined relationship among the water parameters is still in the study and discussion, particularly in estuarine regions of dynamic changes (Zhang and Blomquist., 2018;Hu et al., 2013;Chen et al., 2011;Zhao et al., 2009). It is expected that more detail results and conclusion could be found in the future study and research.