稿件被Geoderma(影响因子1.898)拒了，建议修改后发表在一个应用性的
期刊上。希望有专家能指点一二，感激不尽。
注：PSD指particle-size distribution土壤粒径分布
本文所指PSD模型是指针对PSD发展的各种回归函数

几个关键的问题(下面的红字)：

1，审稿人2指出我没有引用的一篇文章（我没有看过），尽管是1997年
发表的，但密切相关。因此他认为我的文章没有新意，不宜发表。估计没有仔细
看文章了。而这篇文章提出的两个自然指数和幂指数模型仍是基于内
插的（有3-1mm）的数据，与我所引的两项中国人Zhu and Fan, 
1999; Cai et al, 2003的研究类似，而本研究是使用分形概念
进行外插,问题不尽相同（是否能使用是下面第二点的问题）。

Reviewer #2 finds the manuscript unacceptable first because the
scientific content is close to a non-quoted paper previously published
by Rousseva,1997.

If the authors have read the paper, then they should not have repeat the

same idea again just using a different database and different empirical

models. This paper offers nothing new and such empirical comparisons

using various models do not offer much insight into the development of

the particle-size ditribution theory

2，编辑和审稿人1提出模型间经验假设间的冲突。我也不大明白不同假设
的模型是否能同时用于描述一条曲线（现象），即统计学中的不同区间的
假设不同（幂指数和非幂指数）。同时使用不同假设的模型真的不行吗？
（用分形概念做外插，用PSD模型做内插）
这一点是本研究的命门了。（救命啊～～，或者就死掉好了）

I am concerned about the use of a fractal approach to extrapolate

data, since a fractal model is associated with PSD described by

power-laws, at least on sub- domains, and most of the PSD models that

you use contradict this power-law assumption.

the author mixes two distinct models to achieve the conversion（审稿人1）

3，审稿人1对这方面的研究不熟悉，但审稿意见很细致。他/她指出
在转换两个不同系统之间的转换应考虑不同质地分类系统的分级特点。
当然他/她也明白我所用的方法与他举出的例子存在方法上的不同，但
仍认为需要有所讨论。

The above premise is sound but the conversion between two different

systems was previously shown to require the consideration of combined

texture triangle diagram and the related separate limits in each system.

 Although there are some differences in fitting the data and in

summarizing the PSD to its mean and standard deviation in the above

references, the problem needs some discussion

4，另一点也可以很明显的看出，审稿人1对这方面研究是不熟悉的。
因为先估计土壤质地分布，并用它作为输入得到其他土壤水力属性的
研究很多。这样做主要是因为测其他属性花费高。但这种常识性的知识
（包括他/她提到的一些其他疑问），我在文章是没有必要详细阐明的。

I have a basic question relative to fitting data to
 a model then using the model in, say, hydraulic
 properties of a soil. Why not use the observed data
 directly? Can you better develop the need for this
 process? Are there examples to use.







附全文：
COMMENTS FROM EDITORS AND REVIEWERS

Dear *** *** ***,



I very much regret to inform you that your manuscript is not recommended

for publication in our journal. This decision is based on the following

expert reviewers, which I hope you will find useful.

Reviewer #2 finds the manuscript unacceptable first because the

scientific content is close to a non-quoted paper previously published

by Rousseva,1997.

I suggest you to read this paper, to quote it of course, and to improve

your own manuscript taking into account this paper and Reviewer #1's

comments, in order to publish your specific study in a more practical

journal.

I would add some personal comments:

1) I consider that good fitting of a soil PSD model does not mean good

estimation of the soil hydraulic properties (cf. your conclusive

sentence). This depends if there exists some sound relationship between

the parameters of the PSD model and the parameters of the hydraulic

properties. This might be the case with the Van Genuchten mode, since an

explicit link is theoretically proposed between the n, m parameters and

a parameterization of the hydraulic properties, but this is not the case

of all the models you tested.

2) I am concerned about the use of a fractal approach to extrapolate

data, since a fractal model is associated with PSD described by

power-laws, at least on sub- domains, and most of the PSD models that

you use contradict this power-law assumption.

3) In a Geoderma paper, we would avoid some basic definition as

Eq.1 for R2, since this is well known by the scientist audience.

Depending on the type of audience you expect if you publish your

research report in a more applied journal as I recommend you, you can

decide to keep this definition.





With my best regards



Dr Edith Perrier

Joint Editor-in-Chief Geoderma









Reviewer #1: Review of GEODER4664,"An investigation of soil

particle-size distribution models for the conversion of soil texture

classification from ISSS and Kachinsky to USDA System" by *** ***



GENERAL COMMENTS

The author fits three sets of China soil data to as many as ten models

and compares the performance of each model with the data. The majority

of the observed data are expressed in the ISSS system of classification

with the upper cutoff limits for the separates in the sand, silt, and

clay placed, respectively, at 2, .02, and .002 mm. The other two sets

have the upper limits for the sand separate placed at 1 mm with mixed

limits for other separates. These two sets are in Kachinsky system of

classifications.



The expressed objective of the manuscript is to convert data from these

three systems to equivalent textures in the USDA system of

classification with the upper cut off limits for the separates in the

sand, silt, and clay placed, respectively, at 2, .05, and .002 mm.



Although the author does not describe the soil texture triangle for

Kachinsky system of classifications, with which this reviewer has no

familiarity, the paper may be reviewed without that information (see below).



Although not specifically stated, the premise of the conversion to the

USDA system of classification is that a model that fits the PSD of one

data system with predetermined separate limits can be used to describe

the PSD of another system with different separate limits. On this basis,

the author follows routine procedures to fit the data to several

previously used models and obtains his results.



The above premise is sound but the conversion between two different

systems was previously shown to require the consideration of combined

texture triangle diagram and the related separate limits in each system.

For example, Shirazi, et al. (2001), also referenced by the author,

showed that texture conversion from the ISSS system to the USDA system

is always possible, but the reverse conversion is limited on the basis

of the equality of the geometric mean and the geometric standard

deviation of the whole sample (Plate 2C). Those authors explored the

problem in more detail in Shirazi and Boersma (Iranian Journal of

Science and technology, Transaction B, Vol. 25:600-708, 2001).



Although there are some differences in fitting the data and in

summarizing the PSD to its mean and standard deviation in the above

references, the problem needs some discussion, particularly with respect

to Kachinsky system of classification. Strictly, there may exit no

conversion between Kachinsky system of classification and the USDA

system because of the missing upper limit in the sand separate in

Kachinsky system of classification. By using the fractal calculation to

extrapolate the sand separate and one of the ten models for the silt and

clay separates, the author mixes two distinct models to achieve the

conversion. It is not clear to this reviewer that the new PSD is a real

sample representing a specific China soil sample. Please explain.



I have a basic question relative to fitting data to a model then using

the model in, say, hydraulic properties of a soil. Why not use the

observed data directly? Can you better develop the need for this

process? Are there examples to use.





SPECIFIC COMMENTS

This reviewer had difficulty understanding the meaning of "%Unchanged"

in Table 3. The author previously stated that the heading implies

"stability" (line 272). Please explain.



The descriptions "Number of cases for PSD…" on Tables 3 to 5 are not

clear. Please make a statement that better explains the connection

between the numbers in the tables and the goodness of fit, or generality

of the model relative to data sets. Also show the connection between

figures and tables.



Why is there different number of models listed in the Tables and in the

figures? Some are ten and others are eight. What do you want to show by

Tables 4(b, &c) and 5 (a & b).



I am not certain if the question on lines 105-107 can be answered

affirmatively.



Please change "attributions" on line 49 to "properties".



Please explain the sentence at the start of line 123.



Please change "For" on line 126 to "because".



Please remove line 171 and begin the sentence with line 172.















Reviewer #2: The author obviously did not read or forget to cite the

paper by Rouseva who has studied carefully the conversion between the

Katchinski, ISSS and USDA system.

Rousseva, S.S., 1997. Data transformations between soil texture schemes.

European Journal of Soil Science 48, 749-758.



If the authors have read the paper, then they should not have repeat the

same idea again just using a different database and different empirical

models. This paper offers nothing new and such empirical comparisons

using various models do not offer much insight into the development of

the particle-size ditribution theory

相关专题：投稿与审稿
转载本文请联系原作者获取授权，同时请注明本文来自上官微科学网博客。
链接地址：https://blog.sciencenet.cn/blog-4398-50563.html

上一篇：从学生人数看学科教育的发展趋势(土壤科学的例子）
下一篇：So will my life go on