Quantitative assessment of parameterization sensitivity and uncertainty in Noah-MP multi-physics ensemble simulations of gross primary productivity across China&rsquo;s terrestrial ecosystem

Lai, Jie; Wang, Anzhi; Liu, Yage; Shen, Lidu; Zhang, Yuan; Diao, Yiwei; Cai, Rongrong; Li, Rongping; Fei, Wenli; Wu, Jiabing

doi:10.5194/egusphere-2026-103

Preprints

https://doi.org/10.5194/egusphere-2026-103

Preprints

29 Jan 2026

| 29 Jan 2026

Quantitative assessment of parameterization sensitivity and uncertainty in Noah-MP multi-physics ensemble simulations of gross primary productivity across China’s terrestrial ecosystem

Jie Lai, Anzhi Wang, Yage Liu, Lidu Shen, Yuan Zhang, Yiwei Diao, Rongrong Cai, Rongping Li, Wenli Fei, and Jiabing Wu

Abstract. Understanding the carbon cycle and its interactions with climate systems requires precise simulation of Gross Primary Productivity (GPP). However, achieving this remains challenging due to the inherent complexity of the models. Current research lacks quantification of how uncertainties in physical process parameterization affect GPP simulation across various ecosystems, and the dominant physical processes goberning GPP variability are pooly identified. To address these issues, this study generated a 48-member Noah Land Surface Model with multi-parameterization options (Noah-MP) ensemble by manipulating key physical parameterization schemes. The model was validated using ChinaFlux tower measurements and Penman-Monteith– Leuning Version 2 data. We employed the Sobol’ total sensitivity index to assess the influence of four key physical processes on GPP: radiation transfer, the soil moisture limitation factor for transpiration (β-factor), turbulence, and runoff generation. Results demonstrate that Noah-MP effectively captured GPP's spatiotemporal patterns in Chinese ecosystems but overestimated GPP in forest and cropland during spring and summer. Sensitivity analysis indicates that the β-factor dominates GPP simulations across most of China, while radiation transfer is the primary driver on the Tibetan Plateau. The main difference between the two radiation transfer schemes lies in whether vegetation gaps fraction are considered. On the Tibetan Plateau, where grasslands and shrublands exhibit large canopy gaps, consider it or not could lead to in substantial differences in simulated radiation and consequently in GPP, making GPP highly sensitive to the choice of radiation scheme. Across ecosystems, water-related factors (β-factor and runoff) mainly affect croplands and savannas, radiation transfer dominates grasslands and shrublands, and turbulence is most influential in forests. There are also distinct seasonal patterns: radiation and turbulence dominate in spring and summer, while radiation and β-factor prevail in autumn and winter, especially in arid regions. Based on systematic performance evaluations and sensitivity analyses, this study proposes optimized Noah-MP model configurations for China's terrestrial ecosystems. The radiation transfer scheme considering the three-dimensional canopy structure (option 1) is recommended for grasslands and shrublands. Our findings offer insights for enhancing GPP simulation accuracy in Noah-MP, thereby improving the model’s ability to represent carbon–water dynamics from regional to continental scales.

Received: 08 Jan 2026 – Discussion started: 29 Jan 2026

Publisher's note: Copernicus Publications remains neutral with regard to jurisdictional claims made in the text, published maps, institutional affiliations, or any other geographical representation in this paper. While Copernicus Publications makes every effort to include appropriate place names, the final responsibility lies with the authors. Views expressed in the text are those of the authors and do not necessarily reflect the views of the publisher.

Download & links

Preprint (PDF, 2534 KB)

Supplement (1158 KB)

Download & links

Jie Lai, Anzhi Wang, Yage Liu, Lidu Shen, Yuan Zhang, Yiwei Diao, Rongrong Cai, Rongping Li, Wenli Fei, and Jiabing Wu

Status: final response (author comments only)

RC1:
'Comment on egusphere-2026-103', Anonymous Referee #1, 20 Mar 2026
This article is based on the multi-parameterization scheme of the Noah-MP land surface model and systematically evaluates the sensitivity and uncertainty of different physical processes on the simulation of gross primary productivity (GPP) in China's terrestrial ecosystems, which has important scientific significance and practical value. However, I suggest that the authors need to make a major revision to this study and that the study be accepted only after a thorough analysis of the sources of error and uncertainty.

Specific:

Although the study proposes the goal of 'identifying key physical processes,' the introduction does not fully explain the specific gaps in existing research on 'GPP sensitivity analysis in China.' The differences and innovations of this study compared to previous work would stand out more if the authors explicitly stated.

Line 185, Figure 2b. The MODIS land use data used for this vegetation distribution appears to be a significant source of simulation error. It is recommended to replace the land use distribution map and rerun the simulation.

Line 232. The uncertainty in the forcing data needs to be discussed. For example, although CMFD v2.0 is a solid dataset, it inevitably has uncertainties, especially in station-sparse regions like the western Tibetan Plateau. The paper doesn't address whether/how these uncertainties might propagate through the model and affect the sensitivity results. A brief discussion would strengthen the conclusions.

Line 298. The 30-times cycle of spin-up may be another source of uncertainty. For conventional meteorological factors, I believe 30 cycles are sufficient for initiation. However, regarding the authors' research objective of GPP, I doubt this will provide enough driving forces for the integration and accumulation of dynamic vegetation mechanisms. I suggest the authors examine the time-varying curves of variables related to vegetation growth (such as aboveground or total dry weight) during the simulation to see if growth and mortality rates have reached a steady state.

4.1.1. It seems the authors' experiments primarily demonstrated reasonably good performance in farmland, with generally poor results in grasslands, and even worse, yielding outcomes like DHF in forests. Given this performance pattern, how did the authors isolate and address this systematic error?

Section 4.1.2. Should all Figure 3 be Figure 4? Meanwhile, the color bars in Figure 4(a-j) are difficult for me to identify. Could the author please create a different plot for the left and right sides? I think the author needs to seriously analyze their results. Especially the qualitative conclusions drawn after line 391; what we see in Figures 4k and l is not simply a matter of slightly higher or lower. The authors' simulations of GPP using Noah-MP showed that, compared to PML, the results were nearly twice or more than twice as high for various land use types, including EBF, DBF, and Cropland, etc. Why?

Line 412. Typo of ”sensitivityof“. At the same time, I would like to give the author a suggestion in this section(4.2): to conduct regional statistics on these influencing variables according to land use type classification. Now, I can see the constraint relationship between the map and land use type by referring to Figure 2 above. However, the author did not provide a single figure in this entire section, which I think is very irresponsible.

I carefully reviewed the supporting information documents provided by the author. I believe the author needs to examine the significant phase differences between their simulation results and the PML results and explain the reasons for these differences. Furthermore, I suggest that although the goal of this study is GPP, all intermediate diagnostic variables need to be validated in various aspects throughout the simulation process, especially against the flux tower data collected by the author. For example, examine the simulated upward radiation, sensible latent heat flux, soil water, and soil temperature. These can reflect the sources of the author's systematic errors and assist in designing new experiments to improve the GPP simulation level. Only after the overall GPP simulation is reliable should the sensitivity of various parameter combinations be discussed.
Citation: https://doi.org/10.5194/egusphere-2026-103-RC1
RC2:
'Comment on egusphere-2026-103', Anonymous Referee #2, 29 May 2026
This study conducts a comprehensive sensitivity analysis of parameterization schemes on Gross Primary Productivity (GPP) simulations based on the Noah-MP land surface model, focusing on four core physical processes across diverse ecosystems in China. The methodological design, including an ensemble simulation framework, model validation against ChinaFlux and PML-V2 datasets on site scale and regional scale, and the adoption of the Sobol’ total sensitivity index, is technically robust. Nevertheless, given the notable shortcomings in research novelty and mechanistic interpretation that limit its suitability for publication in this journal, I regret that I cannot recommend the acceptance of this manuscript. The detailed comments are provided below.
Major Comments
Numerous studies have been conducted on the sensitivity analysis of Noah-MP parameterization schemes. This manuscript lacks distinct innovation and highlights, and most of the simulation results are predictable and conventional. Furthermore, the physical mechanisms and underlying causes responsible for the model evaluation results are not sufficiently discussed and elaborated throughout the manuscript.

Prior to determining the physical processes for uncertainty analysis, a comprehensive introduction to all physical processes incorporated in the Noah-MP land surface model should be provided. The current manuscript only explains the reasons for selecting the four targeted physical processes, while neglecting to illustrate why other physical processes are excluded from the analysis. For instance, it is necessary to clarify why the uncertainty associated with the parameterization schemes for surface resistance to evaporation does not need to be considered in this study.

Minor Comments
Line 21: Correct the typo “goberning” to “governing”.

Lines 28–29: Quantify the degree of “overestimation” with specific statistical metrics, such as bias or percentage error relative to observational data.

Lines 108–109: The cited references here should focus on the original development and improvement of the Noah-MP model. Representative studies including Niu et al. (2011) and He et al. (2025) are recommended for supplementation.

Line 109: Replace the word “conficting” with “different”, as diverse parameterization schemes do not necessarily conflict with one another.

Lines 125–127: Clearly specify the variable whose simulation uncertainty is discussed in this section.

Lines 224–227: Revise and clarify the logical relationship between the two sentences: when the dynamic vegetation option is activated in Noah-MP, only the Ball-Berry scheme can be adopted.

Line 237: Correct the redundant double punctuation error “..”.

Lines 275 and 281: Unify the writing format of “ERA5-Land” throughout the manuscript to maintain consistency.

Lines 298–301: Please explain the rationale behind the adopted spin-up strategy. Additionally, clarify why the cyclic simulation starts from the year 2000, rather than running a continuous simulation from 1999 during the final spin-up cycle.

Equations (3) and (4): Define the variable 𝜎_f before presenting the relevant formulas.

Line 360: Add in-depth discussion to interpret the underlying reasons for the low TSS values obtained in the results.

Section 4.1.2: Correct the figure citation error; the referenced Figure 3 should be revised to Figure 4.

Line 379: Modify the inappropriate expression. Model simulation outputs should not be referred to as “dataset” in this context.

Section 4.2: This section lacks a detailed description of the seasonal variation characteristics illustrated in Figure 6. Additionally, Figure 7 also summarizes the seasonal variations in the sensitivity of different parameterization schemes. To avoid redundant presentation, it is recommended to retain only one set of the seasonal variation figures and delete the duplicate content.

Line 496: Correct the redundant double punctuation error “..”.

Line 683: Correct the typo “uncertaintie” to “uncertainties”.
Citation: https://doi.org/10.5194/egusphere-2026-103-RC2

Jie Lai, Anzhi Wang, Yage Liu, Lidu Shen, Yuan Zhang, Yiwei Diao, Rongrong Cai, Rongping Li, Wenli Fei, and Jiabing Wu

Supplement

https://doi.org/10.5194/egusphere-2026-103-supplement

Jie Lai, Anzhi Wang, Yage Liu, Lidu Shen, Yuan Zhang, Yiwei Diao, Rongrong Cai, Rongping Li, Wenli Fei, and Jiabing Wu

Viewed

Total article views: 1,660 (including HTML, PDF, and XML)

HTML	PDF	XML	Total	Supplement	BibTeX	EndNote
940	621	99	1,660	154	92	154

HTML: 940
PDF: 621
XML: 99
Total: 1,660
Supplement: 154
BibTeX: 92
EndNote: 154

Views and downloads (calculated since 29 Jan 2026)

Month	HTML	PDF	XML	Total
Jan 2026	235	75	20	330
Feb 2026	285	189	44	518
Mar 2026	305	285	27	617
Apr 2026	68	37	4	109
May 2026	42	30	4	76
Jun 2026	5	5	0	10

Cumulative views and downloads (calculated since 29 Jan 2026)

Month	HTML	PDF	XML	Total
Jan 2026	235	75	20	330
Feb 2026	285	189	44	518
Mar 2026	305	285	27	617
Apr 2026	68	37	4	109
May 2026	42	30	4	76
Jun 2026	5	5	0	10

Viewed (geographical distribution)

Total article views: 1,654 (including HTML, PDF, and XML) Thereof 1,654 with geography defined and 0 with unknown origin.

Country	#	Views	%

Latest update: 25 Jun 2026

Short summary

This study evaluated Noah-MP performance in simulating gross primary productivity (GPP) across China and analyzed the sensitivity of key parameterization schemes. The modified two-stream radiation scheme (RAD01) shows superior performance, especially in grassland and shrubland ecosystems, while the BTR03 β-factor performs better in croplands. Surface exchange and runoff schemes systematically overestimate GPP, indicating structural biases in energy–carbon and hydro–vegetation coupling.


Total:	0
HTML:	0
PDF:	0
XML:	0