the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
| 07 Jul 2026
The Cooling Efficiency Factor Index (CEFI): A New Satellite-Based Dataset for Research and Operational Monitoring of Land Surface Processes
Abstract. The cooling efficiency of the land surface, i.e., its ability to dissipate absorbed radiation and moderate temperature rise, is reflected in its apparent heat capacity, a property that varies throughout the day in response to the relative intensities of sensible and latent heat fluxes. Under clear-sky conditions, the daytime increase in apparent heat capacity can be reliably estimated using geostationary satellite data and used to derive a new dataset called Cooling Efficiency Factor Index (CEFI). This index quantifies land surface energy dissipation through turbulent and ecohydrological processes from 2005 to near real time at a spatial resolution of 5 km. The spatial distribution of the CEFI dataset is primarily determined by land cover, water availability, surface roughness, and wind speed. Its temporal variability can be exploited to derive proxies for variables and processes that are otherwise difficult to observe, especially in real time, such as evapotranspiration and wind speed anomalies. Accordingly, the CEFI dataset can serve as an indicator of vegetation drought stress, the condition in which plants close their stomata due to soil water limitation and high atmospheric water demand, as well as vegetation productivity. It can also detect flash droughts and support improved estimation of fire risk in natural ecosystems, crop production losses in agricultural areas, and dust formation in desert regions. In addition, the dataset can be used to quantify the cooling efficiency of urban areas. This paper provides access to a publicly available CEFI dataset updated in near real time. Given its broad range of applications, the dataset can be used for both research and operational monitoring, to constrain poorly observed processes in dynamical models, and as an additional predictor or predictand in machine learning applications.
- Preprint
(4281 KB) - Metadata XML
- BibTeX
- EndNote
Status: final response (author comments only)
- RC1: 'Comment on essd-2026-296', Anonymous Referee #1, 23 Aug 2026
-
RC2: 'Comment on essd-2026-296', Anonymous Referee #2, 11 Sep 2026
The manuscript ‘The Cooling Efficiency Factor Index (CEFI): A New Satellite-Based Dataset for Research and Operational Monitoring of Land Surface Processes’ by Zampieri et al. presents a new satellite-based dataset of CEFI, which aims to quantify the lands capacity to dissipate heat. Although the authors show strong relations between the CEFI and several biophysical variables I believe several aspects of the manuscript should be improved before publication. Please find my comments below.
Major comments:
- Both the concept (what is CEFI exactly) and goal (why do we need it) of the dataset is lacking throughout the introduction. This makes it unclear what the CEFI adds to the existing drought products. At the same time, the first few paragraphs (L347-381) of the discussions describe this rather clear. Moving these lines to the introduction could provide a scientific and societal relevance that is currently missing. Also, could the authors mention how the CEFI data can be used and maybe give examples of potential applications? Could we use it as a drought index? Or more as a way to study land-atmosphere interactions?
- The derivation of the CEFI in ‘2. Theory’ is not clear without reading Zampieri et al. (2025). I understand that the authors do not want to repeat the earlier paper, but the derivation could be refined. For example:
- Equation 2 On one side there is a delta and on the other side the integral. Maybe make mathematically consistent. Also, it is not clear to me why the cs term is brought to the other side of the equation, and put back in the next step.
- Equation 6: Maybe the authors could shortly explain how the linear approximation follows from cs = ca – ec.
- How is ΔTs exactly defined and calculated? In L74 the authors mention ‘two different times of the day, after the sunrise and before the sunset’, in L111 ‘monotonically increasing LST are used for linear regression’. Does this mean ΔTs is calculated for each time step as the difference between the LST and LST at sunset, as long as LST is monotonically increasing? Also, is this the CEF calculated based on daily values and then averaged over the month?
- Could the authors comment on the uncertainty introduced by the satellite/model products used to compute Rn, and on the independence of these inputs from the variables against which CEFI is later validated? As I understand the methodology, Rn = (1−α)·RS↓ + RL↓ − RL↑, where:
- LW↓ is calculated from an LSASAF product
- LW↑ is calculated using Stefan-Boldzmann with LST (LSASAF) and emissivity (LSASAF)
- SW↓ is calculated from an LSAFAF product
- SW↑ is calculated using the albedo (LSASAF) and SW↓.
- This makes me wonder two things: 1) LST is both used in the Rn and the LST to derive the slope of CEFI. This means errors in LST appear into both Rn and LST itself. Could the authors comment on how this coupling might affect the apparent near-linearity of the ca vs. ΔLST relationship? 2) Several Rn inputs are not directly measured, but are themselves model- or land-cover-derived and may not be independent of the variables used later to validate CEFI. For example, the emissivity is calculated based on look-up-tables which are derived from NDVI. In addition, the downwelling longwave radiation is derived from meteorological forecast fields. Could the authors clarify whether these dependencies were considered and how it affects the results?
- In the results, the authors present that CEFI correlates with several biophysical variables, including soil moisture, NDVI, Wind speed. From the CEFI calculation this makes sense, as both sensible and latent heat flux are included, which depend on things like water availability, vegetation type, atmospheric conditions etc. These correlations seem highly variable in space and time. So how can a user interpret whether a certain CEFI anomaly is due to vegetation stress, or land cover change, or wind conditions etc?
- Did the authors consider looking at anomalies in CEFI? As the CEFI value is highly variable in space and time, looking at (relative) anomalies would provide insight into whether CEFI is different from ‘normal’ conditions. Maybe this would provide more information on drought stress than the actual value?
Minor comments:
- In Figure 1: adding the variable names in the axis labels rather than in the title would improve readability. Also, maybe the exact definition of ΔTs can be added to the figure (i.e. the temperature differences between which hours).
- L103: how do the authors match the different temporal resolutions of the different datasets?
- L112: ‘rejecting regressions with r2 <0.7’. Why were these rejected? How many where rejected based on this criterium? Why do some days have a low r2?
- Line 140-145: if the results are not shown in this paper or elsewhere, adding the information on the PDP does not add anything here.
- L177: NDVI is also strongly dependent on the precipitation regime, right? So how are these impacts separated?
- Figure 3: over which period is the CV calculated? For each month separately?
- Figure 4: it is not clear to me what is meant by ‘mean correlation’ or ‘max correlation’. Does this mean the correlation is calculated per year, for example, and then take maximum value over the years?
- Figure 6: why was 2006 removed? How do you determine it is an outlier? What is the correlation when this outlier would be included? Also, why were other countries not included?
- Add a unit and variable description to the NetCDF files. It would also be useful to add a readme file to the folder.
Citation: https://doi.org/10.5194/essd-2026-296-RC2 -
EC1: 'Comment on essd-2026-296', Tobias Gerken, 17 Sep 2026
Dear authors.
You have now received two reviews with substantial constructive comments including about some of the assumptions that go into CEFI. I am recommending for you to carefully review the comments and to address these in detail in your response and updated manuscript.
Best regards,
Tobias Gerken
Citation: https://doi.org/10.5194/essd-2026-296-EC1
Data sets
Cooling Efficiency Factor Index (CEFI) M. Zampieri and A. Cescatti https://doi.org/10.2905/JRC.401M4D8
Viewed
| HTML | XML | Total | BibTeX | EndNote | |
|---|---|---|---|---|---|
| 307 | 147 | 107 | 561 | 125 | 105 |
- HTML: 307
- PDF: 147
- XML: 107
- Total: 561
- BibTeX: 125
- EndNote: 105
Viewed (geographical distribution)
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1
This review concerns the manuscript entitled “The Cooling Efficiency Factor Index (CEFI): A New Satellite-Based Dataset for Research and Operational Monitoring of Land Surface Processes”. The authors produced a high-spatiotemporal-resolution cooling efficiency index with diurnal variation at a spatial resolution of 5 km for the period 2005–2025, based on geostationary satellite data. The authors find clear relationships between their index and the commonly used meteorological drought index, SPI, as well as with variations in vegetation and soil moisture. Several case studies are presented to demonstrate the integrative ability of the cooling efficiency index to represent land-surface properties at the large scale. I think the manuscript has a clear structure, with easily understandable text and figures. However, I have several comments to be addressed before I can recommend the manuscript for publication in Earth System Science Data.
Major points:
CEFI, the cooling efficiency index, is not purely a consequence of water availability; it also reflects the influence of energy availability that controls nonlinearly on the evaporation. I think this may explain why CEFI is decoupled from NDVI or evaporative fraction in some regions, such as the tropics and drylands (Fig. 5d,g). This decoupling may be particularly relevant when ecosystems experience stress that is not yet extreme and downregulates ET, or when land–atmosphere coupling strengths are changing. I suggest that the authors examine more deeply the joint roles of energy and water variability when interpreting the ecological and biophysical meaning of CEFI dynamics. One simple metric combining the energy and water controls on ET: correlation (ET, soil moisture) minus correlation (ET, net radiation), should be similarly valuable as evaporative fraction but more representing the land surface status, for the authors to have a better understanding of the derived cooling efficiency index.
I do not think the authors clearly explain the advantages of CEFI in the introduction or in the results sections, also when providing showcases regarding its applications. The authors may wish to clarify explicitly the advantages of CEFI relative to commonly used indices such as SPI in the introduction: does CEFI provide a more direct measure of land-surface cooling that goes beyond the atmospheric dryness conditions assessed by SPI? Does it provide a more accurate, near-real-time monitoring approach that could be applied in agricultural or urban studies? Does it lessen the need for ET observations from eddy-covariance measurements or ET and evaporative fraction derived from the satellite LST due to their model uncertainties (see ensemble ET models such as compiled in the OpenET project which also monitor ecosystem stress near-real time: https://etdata.org/methods/)?
For the applications presented in Figures 6 and 8, I would also expect to see comparisons between CEFI and other drought indices in relation to maize production and urban green space, respectively. In addition, the annual-scale maize-production case study may not adequately demonstrate the unique value of CEFI. An analysis at a finer spatial and temporal resolution would be more informative.
For the dust-variation case study, can the authors explain how much of the dust variability is influenced by wind and how much is uniquely associated with cooling efficiency? What processes might link cooling efficiency to dust and air pollution? The authors may wish to control for wind speed when examining the relationship between CEFI and dust. It would also be useful to investigate how local cooling effects, as represented by CEFI, and external wind systems (incl. upwind influences and wind direction) jointly regulate dust formation.
Minor points:
In the introduction, it would be helpful to clarify the advances of this study relative to the previous literature Zampieri et al., 2025.
Li et al., 2023 that is cited in the introduction is missing from the reference list.