Reported validation accuracy of remote sensing estimates of grassland above-ground biomass: a systematic review and meta-analysis

This deposit contains the study protocol, the data-extraction codebook and the extraction template for a systematic review with meta-analysis, in English and in Chinese. The review synthesises the coefficient of determination reported for remotely sensed estimates of grassland above-ground biomass and tests whether reported accuracy is systematically associated with extractable study-design features: validation design, model class, observation platform, sensor, spatial resolution, publication year, region and grassland type. The unit of analysis is 2,878 de-duplicated journal and conference records with bibliographic data and abstracts already in hand. Full texts are used to verify values when obtainable; failure to obtain a full text is not an exclusion criterion and such records are flagged as abstract-level only. The effect size is the logit transformation of R². Inference uses equal weights, study-clustered robust standard errors and a study-level bootstrap; no closed-form sampling variance is assumed. No minimum number of effect sizes, no minimum number of studies in any category, no deadline and no downgrade clause are specified. Searches of Web of Science Core Collection, CNKI and Wanfang were completed and frozen on 2 November 2026: 3,090 records exported, and 2,878 records entered the screening frame after de-duplication. Recall was validated against 23 independent benchmark studies and rose from 16/23 to 23/23 after an accuracy-term block was removed from the strategy. Title-abstract screening has not begun, no screening decision has been recorded, and no pooled estimate has been computed. The protocol was preregistered on the Open Science Framework on 4 November 2026 (registration DOI 10.17605/OSF.IO/U37KF, https://osf.io/u37kf). The three English files are identical to the files attached to that registration; the three Chinese files are translations prepared for the author's research group.