PlumX Metrics
Embed PlumX Metrics

Radiation Partitioning in a Cloud-Rich Tropical Mountain Rain Forest of the S-Ecuadorian Andes for Use in Plot-Based Land Surface Modelling

SSRN, ISSN: 1556-5068
2024
  • 0
    Citations
  • 32
    Usage
  • 0
    Captures
  • 0
    Mentions
  • 0
    Social Media
Metric Options:   Counts1 Year3 Year

Metrics Details

  • Usage
    32

Article Description

The understanding of how incoming shortwave radiation is divided into direct and diffuse components is crucial for comprehending ecosystem energy flows and carbon storage. Accurate partitioning functions are necessary for modeling studies, often carried out with land surface models (LSM) coupled with climate models. However, these functions largely rely on regional cloud and aerosol conditions. Data for developing semi-empirical partitioning functions are mostly available for the mid-latitudes, and there is a lack of knowledge about their performance in the tropics. This particularly holds for tropical forest ecosystems in the high Andes, where ground-based data on cloudiness and irradiation components are typically scarce. As a result, it is uncertain whether commonly applied functions can be utilized in regional LSMs without adaptation. This study evaluates a unique dataset of shortwave radiation components for a tropical mountain rainforest (MRF) in southern Ecuador. It also derives and tests a locally adapted shortwave radiation partitioning function based on machine learning, compares its performance with other semi-empirical functions, and evaluates the impact of partitioning functions on simulated outgoing shortwave radiation with an LSM. The findings reveal a consistently high cloud cover and prevailing diffuse radiation fluxes, especially during the rainy season. A higher proportion of direct irradiation is observed in the shorter, relatively dry season (Veranillo del Niño). The locally derived partitioning function performs well and outperforms other commonly used functions. Consequently, simulations of outgoing shortwave radiation using a plot-based LSM demonstrate improved performance when implementing the derived partitioning function. These results emphasize the high importance of locally and regionally adapted radiation partitioning functions based on machine learning and a more comprehensive feature set. The adapted LSM will be further utilized for heat flux and carbon sequestration studies.

Bibliographic Details

Paulina Grigusova; Oliver Limberger; Jörg Bendix; Charuta Murkute; Franz Pucha; Katja Trachte; Victor Hugo González-Jaramillo; Andreas Fries; David Windhorst; Lutz Breuer; Mateus Dantas de Paula; Thomas Hickler

Elsevier BV

Multidisciplinary; Land Surface Model; ecuador; Radiation; partitioning; Machine Learning; tropical mountain rain forest

Provide Feedback

Have ideas for a new metric? Would you like to see something else here?Let us know