Overview

The work by Seitz and colleagues presents a new scheme for the senescence of leaves in grassland areas. In particular, the presented scheme uses three main drivers: temperature, soil moisture, and day length. This approach led to an improved representation of leaves within the QUINCY model, where it is applied, highlighting the need for decoupling leaf senescence from growing season triggers. Moreover, the presented methodology exhibits promising results on a set of sites within different climatic zones, showing a possible development for other Land Surface Models.

However, some state-of-the-art land surface models (LSMs) already account for the same set of drivers within their senescence parameterizations, such as the Community Land Model (Lawrence et al., 2019) and LPJ-GUESS (Smith et al., 2014). The authors should extend their discussion on the differences, similarities, and complementarity of their approach compared to the ones implemented in other LSMs. Besides this main point, some typos should be addressed before the manuscript is ready for publication.

Specific comments

Figure 1c: Is the light function symmetrical with respect to the equator?

Line 222: “[…] only of the factors is limiting. […]” should be “[…] only one of the factors is limiting. […]”

Lines 228-229: “The results of our sensitivity test reveal that 10% or a shedding of all leaves over 10 days at the highest shedding rate (fshed,max = 0.1) is a more suitable value for our model.”  Are these sensitivity tests evaluated against observations? If yes, it would be worth mentioning here.

Lines 229-230: “Finally, we have added an additional condition to the growing season in the dynamic model so that an accumulated snow depth of > 2cm stops the growing season.”  Is a snow depth of 2cm creating an immediate shedding of all leaves? Is it applied throughout the year or only during the fall season?

Table 2: add also the reference of Nair et al. (2024) for the ES-LMa site.

Caption Figure 3: describes in detail the level of significance given by the letters a and b.

Figure 3: ES-LMa site provides observed data of GPP from 2003 to 2006 (4 years) plus 2015-2018 (4 years) as reported in table 2 and Figure 2. Which are the 7 years used by the authors in figure 3?

Caption Figure 4: “ecosystem C (a), vegetation C (b) and soil organic carbon (SOC; c)”. The panel letter could be placed before the variable names, such as “(a) ecosystem C, (b) vegetation C, and (c) soil organic carbon (SOC)” to improve readability.

Line 372: “similar model fits (Table 5,” closing bracket is missing.

Line 602: NDdormancy should be consistent with equation A1 and written as NDdormance.

Line 616: GPPtlabile should be GPPlabile

Figure B1: Is it possible to distinguish the 37 sites used for evaluating the EOS timing as well? Perhaps report those sites with a square instead of a circle on the map

The detailed comments are provided in the attached PDF file