Workflow and programming tool to derive empirical models of primary productivity. Study case in Doñana National Park wetlands : master thesis

Fuentes Monjaraz, Mario Alberto (author), El Serafy, Ghada (mentor), Sperna Weiland, Frederiek (co-mentor), Spinosa, Anna (co-mentor), Weerts, Albrecht (co-mentor)

Multiple international agreements have emphasised the use of monitoring activities in ecosystems for better conservation of biodiversity and sustainable use of natural resources. Even though the monitoring activities have improved with Earth Observation products, there is still a lack of understanding and research in monitoring ecosystem functions. Primary Productivity (PP) is a process that supports most of the ecosystem functions, and its estimations is a step forward for a better understanding and monitoring of this component of biodiversity. This thesis proposes a workflow and a programming tool to produce empirical models of primary productivity and map Gross Primary Productivity (GPP) and Net Primary Productivity (NPP) in ecosystems at a local scale with high-resolution data. The workflow follows previous research on remote sensing techniques for primary productivity modelling. Specifically, the empirical method described by Cai et al. (2021) is implemented to derive GPP empirical models. A methodology is proposed to upscale the model and compute annual GPP predictions. The annual GPP predictions are finally used to approximate the annual NPP with the Carbon Use Efficiency (CUE) conceptual parameter of the ecosystems reported in previous research. The workflow is applied in a wetland ecosystem in Doñana National Park. Data scarcity in the study area limited the application of the workflow. From the original workflow targets, only a GPP model could be formulated for an 8- month period. The model predicted GPP with high performance deriving MAE = 0.52 gC m-2 day-1, RMSE = 0.63 gC m-2 day-1 and R2 = 0.93. The average 8-month GPP in the monitored ecosystem was estimated at 770.20 gC m-2. The research insights are relevant for future research on monitoring primary productivity in wetland ecosystems with empirical approaches. Especially, the use of the Red-edge Chlorophyll Index formulated with high spectral and high spatial resolution Sentinel-2 data is highlighted for further research on remote sensing modelling of primary productivity.

• Language: English
• Keywords: civil engineering, master thesis, primary productivity, remote sensing, eddy covariance, ecosystem functioning, wetlands, sustainability.
• Work type: Master's thesis/paper
• Typology: 2.09 - Master's Thesis
• Organization: FGG - Faculty of Civil and Geodetic Engineering
• Place of publishing: Ljubljana
• Publisher: [M. A. Fuentes Monjaraz]
• Year: 2022
• Number of pages: XII, 112 str.
• PID: 20.500.12556/RUL-141727
• UDC: 574(043.3)
• COBISS.SI-ID: 123920899
• Publication date in RUL: 06.10.2022

Secondary language

• Language: Slovenian
• Title: Metodologija in modeliranje empiričnega določanja primarne produkcije: primer nacionalnega parka Doñana : magistrsko delo

Abstract

Številne mednarodne zaveze so poudarile pomembnost spremljanja ekoloških razmer različnih ekosistemov z namenom izboljšanja biodiverzitete in trajnostne rabe naravnih virov. Kljub temu, da so se metodologije opazovanja izboljšale, je še vedno precej neznank povezanih z opazovanji in razumevanjem ekosistemskih storitev. Primarna produkcija (PP) je proces, ki podpira večino ekosistemskih storitev, in ocena primarne produkcije je korak k izboljšanju biodiverzitete. V nalogi je predstavljen razvoj metodologije in modeliranje empiričnega modela za primerno produkcijo in karte t.i. bruto primerna produkcije (GPP) in neto primarne produkcije (NPP) v ekosistemih na lokalnem nivoju. Metodologija temelji na zadnjih dognanjih s področja daljinskega zaznavanja. Empirična metoda, ki so jo opisali Cai et al. (2021) je uporabljena kot osnova za določitev GPP modela. Predstavljena je metodologija za izračun letnih vrednosti GPP na večjem območju. Letne vrednosti GPP so uporabljene za oceno NPP vrednosti z upoštevanjem učinkovitosti ogljika (CUE), ki je bil kot parameter upoštevan tudi v predhodnih raziskavah. Metodologija je testirana na primeru mokrišč v nacionalnem parku Doñana. Pokazale so se določene omejitve pri uporabi metodologije predvsem z vidika razpoložljivosti podatkov. Določen je bil GPP za 8-mesečno obdobje. Učinkovitost modela je bila visoka z MAE = 0,52 gC m-2 dan-1, RMSE = 0,63 gC m-2 dan-1 in R2 = 0,93. Povprečen 8-mesečni GPP je bil ocenjen na 770,20 gC m-2. Zaključki raziskave so pomembni z vidika nadaljnjega razvoja določanja primarne produkcije v mokriščih z uporabo empiričnih metod. Posebej uporaba t.i. Red-edge indeksa v kombinaciji s podatki Sentinel-2 se je izkazala kot primerna za nadaljnjo uporabo v podobnih raziskavah.

• Keywords: gradbeništvo, magistrska dela, VOI, primerna produkcija, daljinsko zaznavanje, metoda turbulentne difuzije, ekosistemske storitve, mokrišča, trajnostni razvoj