https://geogeeklab.github.io/Research notes on GIS, remote sensing, GeoAI, visualization, and geographic observation.enhttps://geogeeklab.github.io/field-notes/limits-of-ai-geospatial-foundation-models/https://geogeeklab.github.io/field-notes/limits-of-ai-geospatial-foundation-models/Invalid DateGeospatial foundation models can learn powerful spatial representations. Their harder test is geographic reasoning: scale, spatial dependence, regional heterogeneity, mechanism, and transfer beyond familiar places.https://geogeeklab.github.io/field-notes/machine-learning-remote-sensing-good-practice/https://geogeeklab.github.io/field-notes/machine-learning-remote-sensing-good-practice/Invalid DateThe hard part is not fitting a model. It is proving that what the model learned survives sparse samples, spatial heterogeneity, weak signals, and geographic transfer.https://geogeeklab.github.io/field-notes/nature-reviewer-skills-450-critique-patterns/https://geogeeklab.github.io/field-notes/nature-reviewer-skills-450-critique-patterns/Invalid DateA review-assistance system built from public peer-review files: not a substitute for judgment, but a structured catalogue of the questions reviewers repeatedly ask.https://geogeeklab.github.io/field-notes/distilling-nature-reviewers-earth-system-science/https://geogeeklab.github.io/field-notes/distilling-nature-reviewers-earth-system-science/Invalid DateA domain-specific extension of reviewer-pattern mining across atmosphere, hydrology, ecology, and remote sensing.https://geogeeklab.github.io/field-notes/nature-remote-sensing-reviewer-skill/https://geogeeklab.github.io/field-notes/nature-remote-sensing-reviewer-skill/Invalid DateA lightweight reviewer skill assembled from public review files, focused on the points where remote-sensing papers most often lose credibility.https://geogeeklab.github.io/field-notes/alphaearth-physical-semantics/https://geogeeklab.github.io/field-notes/alphaearth-physical-semantics/Invalid DateA new study asks whether AlphaEarth’s latent dimensions correspond to interpretable land-surface properties rather than merely useful predictive features.https://geogeeklab.github.io/field-notes/writing-causal-claims-ecology/https://geogeeklab.github.io/field-notes/writing-causal-claims-ecology/Invalid DateCausal language is earned by a question, a defensible causal structure, an identification strategy, and sensitivity analysis—not by significance alone.https://geogeeklab.github.io/field-notes/greenness-is-not-productivity/https://geogeeklab.github.io/field-notes/greenness-is-not-productivity/Invalid DateGreenness, vegetation cover, and productivity often move together—but global evidence shows that they diverge across large parts of the vegetated world.https://geogeeklab.github.io/field-notes/chlorophyll-is-not-biomass/https://geogeeklab.github.io/field-notes/chlorophyll-is-not-biomass/Invalid DateChlorophyll-a and phytoplankton biomass can decouple, vary for different physiological reasons, and in some regions even move in opposite directions.https://geogeeklab.github.io/field-notes/global-cropland-dynamics-2015-2024/https://geogeeklab.github.io/field-notes/global-cropland-dynamics-2015-2024/Invalid DateA new Landsat-based annual cropland dataset shows where farmland expanded, where it contracted, and how trade, conflict, drought, and land conversion left different regional signatures.https://geogeeklab.github.io/field-notes/proxy-of-a-proxy-remote-sensing/https://geogeeklab.github.io/field-notes/proxy-of-a-proxy-remote-sensing/Invalid DateHigh retrieval accuracy does not guarantee a valid environmental conclusion. Remote sensing often estimates a measurable variable first and interprets the target object only through a second proxy relation.https://geogeeklab.github.io/field-notes/qiusheng-wu-open-source-geospatial-learning/https://geogeeklab.github.io/field-notes/qiusheng-wu-open-source-geospatial-learning/Invalid DateA personal note on the educator whose open-source teaching made GIS and remote sensing feel learnable by building, not by memorising software menus.https://geogeeklab.github.io/field-notes/chl-a-retrieval-versus-biomass/https://geogeeklab.github.io/field-notes/chl-a-retrieval-versus-biomass/Invalid DateThe satellite product may be chlorophyll-a; the paper may discuss biomass. The missing step between them is an ecological representation relation, not part of the retrieval itself.https://geogeeklab.github.io/field-notes/autocorrelation-and-false-trends/https://geogeeklab.github.io/field-notes/autocorrelation-and-false-trends/Invalid DateLong time series contain memory. If that dependence is ignored, apparent accelerations, breakpoints, and regime shifts can look more certain than the evidence allows.https://geogeeklab.github.io/field-notes/earth-system-maxwell-equations/https://geogeeklab.github.io/field-notes/earth-system-maxwell-equations/Invalid DateA Wuhan University perspective proposes a planetary laboratory in which global satellite observations and AI are used to search for compact, interpretable laws of Earth-system behaviour.https://geogeeklab.github.io/field-notes/temporal-consistency-long-remote-sensing-records/https://geogeeklab.github.io/field-notes/temporal-consistency-long-remote-sensing-records/Invalid DateA decades-long satellite series is rarely one instrument watching continuously. Sensor changes, orbit drift, calibration, fusion, and processing can create trends that look environmental.https://geogeeklab.github.io/field-notes/when-nature-disagrees-remote-sensing-observation/https://geogeeklab.github.io/field-notes/when-nature-disagrees-remote-sensing-observation/Invalid DateTwo apparently conflicting studies of Amazon seasonality expose two different failures of remote sensing: seeing a signal created by geometry, and failing to sample enough of the phenomenon.https://geogeeklab.github.io/field-notes/causal-shapley-is-not-causal-identification/https://geogeeklab.github.io/field-notes/causal-shapley-is-not-causal-identification/Invalid DateAdding a causal graph to Shapley attribution can change the decomposition, but it does not by itself identify a causal effect from observational data.https://geogeeklab.github.io/field-notes/three-boundaries-of-shap-in-nature-papers/https://geogeeklab.github.io/field-notes/three-boundaries-of-shap-in-nature-papers/Invalid DateThree Nature-family papers show that Shapley values can play very different epistemic roles depending on the surrounding evidence.https://geogeeklab.github.io/field-notes/shap-interpretation-boundaries/https://geogeeklab.github.io/field-notes/shap-interpretation-boundaries/Invalid DateThe question is not whether SHAP is “misused” in the abstract, but which class of claim the evidence actually supports.https://geogeeklab.github.io/field-notes/extrapolation-risk-global-scale-studies/https://geogeeklab.github.io/field-notes/extrapolation-risk-global-scale-studies/Invalid DateA global map can look continuous while its evidence remains local. Three questions matter: the input data, the validation design, and the boundary of extrapolation.https://geogeeklab.github.io/field-notes/reassessing-global-tree-restoration-potential/https://geogeeklab.github.io/field-notes/reassessing-global-tree-restoration-potential/Invalid DateThe influential 2019 Science estimate of global restoration potential became a case study in how basemaps, validation, and extrapolation shape policy-scale numbers.https://geogeeklab.github.io/field-notes/alphaearth-agriculture-benchmark/https://geogeeklab.github.io/field-notes/alphaearth-agriculture-benchmark/Invalid DateAlphaEarth is competitive under local training, but its advantage weakens as agricultural tasks demand cross-region transfer, fine temporal sensitivity, and physical interpretation.