Remote-C: a (M)MRV technology to ensure decarbonization in the Italian agri-food system through C removals
Bertola, Marta; Croci, Michele; Ragazzi, Manuele; Marcone, Andrea; Ofori-Karikari, Kwasi Appiah-Gyimah; Cremonesi, Lorenzo; Amaducci, Stefano; Ferrarini, Andrea
- Publisher
- Zenodo
- Language
- en
Abstract
Developing scalable and locally tailored regenerative farming solutions is essential for fostering resilient agroecosystems. The TomaTO-C living lab, funded by the ClienFarms and Farms4Climate projects, co-designs and implements Operational Protocols for Regenerative Agriculture (OPRA) while enrolling commercial farms into a carbon farming (CF) program. In the MARVIC project, a context-specific hybrid Measuring, Monitoring, Reporting, and Verification ((M)MRV) scheme, called Remote-C, has been developed to measure, monitor, and verify CRCF-aligned carbon removal certificates at farm and program level. Remote-C integrates digital sensing technologies, FMIS and IACS/LPIS data, smart soil sampling, advanced soil C modeling, and remote sensing (RS) products into an Operational Processing Chain (OPC). Remote-C addresses key challenges such as data assimilation, semi-automatic data entry, scalability and cost-accuracy of estimates SOC sampling and analysis costs, and uncertainty propagation. It performs ex-ante parcel-specific regional baselines through model ensemble of RS-based SOC retrieval and Roth-C model simulations. Initial soil sampling uses either RS-based stratified strategies or on-the-go SOC stock accounting methods based on proximal sensors, with CN analyzers and/or MIR spectrometry analysis. During the ex-post phase, Remote-C allow monitoring of plot-level inputs for Roth-C model, including meteo ERA5 data, crop and management records (from IACS, FMIS or RS products), and RS detection of agricultural operations (e.g., crop phenology, residue presence, tillage events). These inputs inform continuous soil cover factor calculations and temporal allocation of plant C inputs to soil. Crop biomass and yields are estimated at 10-meter resolution using a Light Use Efficiency (LUE) model that integrates multiple satellite imageries and meteo data. Currently being included in a digital platform as part of the AGRITECH flagship project, Remote-C ensures transparency of certificates, supports in-setting programs, and enables agrifood companies to meet SBTi net-zero targets. By addressing technical and economic barriers, Remote-C aims to ultimately accelerate the adoption of sustainable CF practices.
Full text
𝑷𝒍𝒂𝒏𝒕 𝑪 𝒊𝒏𝒑𝒖𝒕 = 𝑺𝑶𝑺 𝑻(𝑪𝒆𝒙𝒖𝒅𝒂𝒕𝒆𝒔 +𝑪𝒓𝒐𝒐𝒕𝒔 +𝑪𝒓𝒆𝒔𝒊𝒅𝒖𝒆𝒔) Bolinder et al. 2007 ACRONYMS: •SOC: Soil Organic Carbon •CRCF: Carbon Removal Certification Framework •GHG LS&R: Greenhouse Gas Protocol Land Sector and Removals •SBTi FLAG : Science Based Targets initiative for Forest, Land and Agriculture Sector •OPC: Operational Processing Chain •RS: Remote Sensing •LPIS: Land Parcel Identification System •FMIS: Farm Management Information System •IACS: Integrated Administration and Control System •SOS & EOS: Start and End of the Season •EOS: End of the Season •AGDB: AboveGround Dry Biomass •LUE: Light Use Efficiency •fAPAR and PAR: fraction Absorbed of Photosynthetically Active Radiation and Photosynthetically Active Radiation •SBSI: Synthetic Bare Soil Index •ATV: all-terrain vehicle •PMU: Pooled Measurement Uncertainty DPM/RPM from Dechow et al. 2019 Stratified systematic random «SMART» sampling •Bayesian Roth-C & LUE models calibration for defined domain and multiple eligible CF practices using Living lab and benchmark sites data •Uncertainty propagation (model+ sampling + input data) across OPC •Inclusion of validated new RS products (e.g. cover crop detection) •Automation of input data assimilation and modelling into a digital platform Residues permanence & cover Configure and enroll Monitor and Measure Report and Verify Field eligibility Eligible field for establishing a dynamic, activity-specific baseline (left) and an enrolled field of the CF project (right) Remote-C: a (M)MRV technology to ensure decarbonization in the Italian agri-food system through C removals Time series reconstruction - S2 gap free Vis (NDVI, EVI2, PPI) and S2 bands Ex-ante baseline Ex-post baseline CF project line SOS POS EOS ERA5 Land Define project DOMAIN (arable land use × pedoclimatic conditions × common crop rotation) 1 Department of Sustainable Crop Production –Università Cattolica del Sacro Cuore, Piacenza, Italy; 2 Remote Sensing and Spatial Analysis Research Center (CRAST), Università Cattolica del Sacro Cuore, 29122 Piacenza, Italy; 3 Centro per l′Innovazione nell’Impiego del Telerilevamento nell’Industria Meccanica per l′Agricoltura di Precisione (CITIMAP), via Castellarino, 12, San Bonico, PC 29122, Italy; 4 Consorzio Agrario Terrepadane, Piacenza, Italy Marta Bertola1, Michele Croci1,2, Manuele Ragazzi1, Andrea Marcone1,3, Kwasi Appiah-Gyimah Ofori-Karikari1, Lorenzo Cremonesi1,4, Stefano Amaducci1,2,3, Andrea Ferrarini1 Motivation and aims II European Carbon Farming Summit 04-06 March 2025 Dublin, Ireland Remote-C is a Tier 3 hybrid (M)MRV technology designed to overcome scalability and cost-accuracy challenges in carbon farming (CF) programs. By leveraging data assimilation and modelling frameworks, the simulation of SOC stocks under baseline and CF scenarios can be conducted at high resolution across multiple scales to measure C removals (CR). Remote-C is designed to configure and support MRV activities of initiatives such as TomaTO-C living lab and the upcoming offsetting & insetting programs that will be following CRCF, GHG LS&R, SBTi FLAG, or Verra VM0042 standards. Remote-C functions as an OPC for a defined project domain in Northern Italy. In the ex-ante phase, it establishes regional standardized baselines by combining RSbased SOC retrieval with Roth-C simulations. On the enrolled fields, Roth-C is initialized using ESM-derived SOC stock data or maps obtained through smart soil sampling. In the ex-post phase, it quantifies parcel-level C removals using a calibrated Roth-C model fed by FMIS/IACS data, remotely sensed activity data (e.g., C-factor, tillage events), and plant C input data. Crop biomass and yields are estimated using LUE model, which assimilates RS data for crop phenology and fAPAR estimation. Methods Cresidues (if incorporated) K-means via silhouette Neyman allocation # sampling per cluster clhs Sampling position Homogenous area (cluster) #sampling Proximal sensors (γ ray, EC, vis-NIR) SOC (%) Clay (%) SOC stock (ESM ton ha-1) ATV + autosampler Tillage events Bare soil period T Exogenous organic matter (EOM) C Plant material C Soil active C /BIO Soil passive C /IOM CO2 fHEOM fDEOM KDEOM KREOM fREOM fDPM fRPM KHEOM KDPM KRPM CO2 Soil slow C /HUM EOM resistant C /EOM EOM decomposable C /EOM Soil metabolic C /DPM Soil structural C /RPM KBIO KHUM = 𝑺𝑶𝑺 𝑬𝑶𝑺 εmax fAPAR PAR TsWs ×××× AGDB (gC m-2) Model validation report LPIS/ IACS and/or FMIS Cover crops (summer, winter) Minimum and strip tillage Crop residues incorporation Eligible “additional” CF practices SOS EOS Crop Phenology IACS and/or FMIS Organic amendments Future implementation LUE model Vr Farm 1 Farm 2 … Farm n Water (Ws) and temperature (Ts) stressors PAR fAPAR = f(NDVI) εmax SBSI S2 bands SOC stock (ton C ha-1) Beginning CF project Pre-CF project reference on 5 years CF Project period 2016 2017 2018 2019 2020 2021 2022 2023 2024 2025 Pre-CF project baseline Ex-post CF project line Ex-ante CF project line Ex-post dynamic baseline Ex-ante dynamic baseline Ex-post C removals of CF project Ex-ante C removals of CF project Sentinel 2 (S2) 𝑺𝑶𝑺 𝑬𝑶𝑺𝒙 T Constraints Lab analysis Prediction model Initialization Verification Ex-ante Ex-post Continuous C-factor 1 0.60.8 0.6 0.8 1 0.9 0.7 0.6 0.6 0.6 0.8 1 0.8 0.6 1 0.60.8 1 Cover crop detection Ypool = Y0e−a,b,c,d,k𝑡 a : Temp Rate Modifier (RM) b : Humidity RM c : Cover RM d : Tillage RM Kpool : decomposition constant rate Roth-C model 3 6 5 4 1 8b 8a 7 9 8c 11 10b 10a Model spin-up & warm-up 11a 11b 12 Vr Vr 2Field boundaries and crop type Vr Vr PMU, model bias, confidence coverage of at least 90% farms for 90% prediction intervals if cost-effective Discrete values Raster maps Organic fertilization Diversified crop rotation T In 5 8+ TVr 13 14 Certificates issue Multiple high resolution soil and land use maps of Emilia Romagna region SOC stock 30 cm Mg ha-1 Estimated AGDB in Feb 23’ with cover crops T Yield (Mg ha-2) To be used for - GHG calculators - Yield monitoring - Precision farming - Leakage calculation
