[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"doc-detail-127744-en":3,"doc-seo-127744-105":30,"detail-sidebar-cat-0-en-105":92},{"code":4,"msg":5,"data":6},0,"success",{"doc_id":7,"user_id":8,"nickname":9,"user_avatar":10,"doc_module":4,"category_id":11,"category_name":12,"doc_title":13,"doc_description":14,"doc_content":15,"file_id":16,"file_url":17,"file_type":18,"file_size":19,"view_count":20,"is_deleted":4,"is_public":20,"is_downloadable":20,"audit_status":20,"page_count":21,"language":22,"language_code":23,"site_id":24,"html_lang":23,"table_of_contents":25,"faqs":26,"seo_title":27,"seo_description":14,"update_tm":28,"read_time":29},127744,962084926284,"Aurora","https://ap-avatar.wpscdn.com/davatar_29158cc5080c5b710cf443261637dec0",8,"Research & Report","Local interpretation of machine learning models in remote sensing with SHAP - the case of global climate constraints on photosynthesis phenology","Data-driven machine-learning models are widely used in remote sensing for retrieving biophysical variables and classifying land cover, yet they often act as “black boxes” whose input–output relationships are difficult to interpret. Weather influences on downscaled sun-induced fluorescence (SIF) have remained unclear. Using SHapley Additive exPlanations (SHAP), the study explains each input variable’s contribution in a weather-SIF regression. SHAP estimates across the globe quantify how air temperature, shortwave radiation, and vapor-pressure-deficit constrain photosynthetically active seasons and clarify limiting factors in tropical and extra-tropical regions.","This is the accepted version of the journal article:  \nDescals, Adrià; Verger, Aleixandre; Yin, Gaofei; [et al.] . «Local interpretation of machine learning models in remote sensing with SHAP : the case of global climate constraints on photosynthesis phenology». International Journal of Remote Sensing, Vol. 44, issue 10 (2023), p. 3160-3173. DOI 10.1080/01431161.2023.2217982  \nThis version is available at [https://ddd.uab.cat/record/287444](https://ddd.uab.cat/record/287444)[ ](https://ddd.uab.cat/record/287444)under the terms of the  license  \n1 TITLE: Local interpretation of machine learning models in remote sensing with  \n2 SHAP: the case of global climate constraints on photosynthesis phenology  \n3  \n4 Adrià Descals 1,2, Aleixandre Verger1,2,3,Gaofei Yin4, Iolanda Filella 1,2, and Josep Peñuelas 1,2  \n5 1CREAF, Cerdanyola del Vallès, Barcelona 08193, Catalonia, Spain  \n6 2CSIC, Global Ecology Unit CREAF‐CSIC‐UAB, Bellaterra, Barcelona 08193, Catalonia, Spain  \n7 3CIDE, CSIC-UV-GV, València 46113, Spain  \n8 4 Faculty of Geosciences and Environmental Engineering, Southwest Jiaotong University, Chengdu 610031, China  \n9  \n10 Abstract  \n11 Data-driven models using machine learning have been widely used in remote sensing  \n12 applications such as the retrieval of biophysical variables and land cover classification. However, 13 these models behave as a ‘black box’, meaning that the relationships between the input and  \n14 predicted variables are hard to interpret. Recent regression models that downscale sun-induced  \n15 fluorescence (SIF) with MODIS and weather variables are an example. The impact of weather  \n16 variables on the predicted SIF in these models is unknown. The explanation of such weather-SIF  \n17 relationships would aid in the understanding of climate-related constraints on photosynthesis  \n18 phenology since SIF is a proxy of gross primary productivity. Here, we used SHapley Additive  \n19 exPlanations (SHAP)–a novel technique based on game theory– for explaining the contribution  \n20 of input variables to the individual predictions in a machine learning model. We explored the  \n21 capabilities of this technique with a weather-SIF model. The regression model predicted ESA- 22 TROPOSIF measurements from ERA5-Land air temperature, shortwave radiation, and vapor- 23 pressure-deficit (VPD) data. The SHAP values of the model were estimated at the start and end  \n24 of the growing season for the entire globe. These values depicted the global constraints of the  \n25 three climate variables on the photosynthetically active season and confirmed existing  \n26 knowledge on the limiting factors of terrestrial photosynthesis with unprecedented spatial detail.  \n27 Radiation was the limiting factor in tropical rainforest and VPD constrained the start and end of  \n28 the growing season in tropical dryland ecosystems. In extra-tropical regions, temperature was  \n29 the main limiting factor during the start of the growing season, but both temperature and  \n30 radiation constrained photosynthesis at the end of the growing season. This technique may help  \n31 future remote sensing studies that require the use of non-interpretable machine-learning  \n32 regression models and explain how input variables contribute to the model prediction in a  \n33 spatiotemporally explicit manner.  \n34 Keywords: SHapley Additive exPlanations, explainable machine learning, local interpretation, 35 sun-induced fluorescence, vegetation phenology, climate constraints, photosynthesis dynamics.  \n36 1. INTRODUCTION  \n37 The field of vegetation phenology has gained attention recently, with the number of publications  \n38 on phenology quintupling in the last two decades (Fu et al., 2020) . The transition between the  \n39 dormant and growing season and the climate factors determining it have been explained globally  \n40 by models employing climate thresholds. Jolly et al. (2005) proposed the growing season index  \n41 (GSI), which is calculated with cut-off func","cbCairKlGlp7g3nj","https://ap.wps.com/l/cbCairKlGlp7g3nj","pdf",1100749,1,24,"English","en",105,"# Abstract\n# Introduction\n## Vegetation phenology and climate thresholds\n## Machine-learning downscaling of sun-induced fluorescence\n## Need for interpretability and local explanations","[{\"question\":\"Why is interpretability important for machine-learning remote sensing models?\",\"answer\":\"Because these models often behave as “black boxes,” making the relationships between inputs and predicted outputs difficult to understand.\"},{\"question\":\"How does SHAP contribute to interpreting the weather-SIF regression model?\",\"answer\":\"SHAP estimates the contribution of each input variable to individual predictions, enabling local interpretation of how inputs drive outcomes.\"},{\"question\":\"Which climate variables were identified as limiting factors, and where?\",\"answer\":\"Radiation limited the photosynthetically active season in tropical rainforest, while vapor-pressure-deficit constrained start and end in tropical drylands; temperature limited the start in extra-tropical regions, and both temperature and radiation constrained the end.\"}]","Local interpretation of machine learning models in remote sensing with SHAP - the case of global climate constraints on photosynthesis phenology | PDF",1785941363,60,{"code":4,"msg":31,"data":32},"ok",{"site_id":24,"language":23,"slug":33,"title":13,"keywords":34,"description":14,"schema_data":35,"social_meta":87,"head_meta":89,"extra_data":91,"updated_unix":28},"local-interpretation-of-machine-learning-models-in-remote-sensing-with-shap-the-case-of-global-climate-constraints-on-photosynthesis-phenology","",{"@graph":36,"@context":86},[37,54,69],{"@type":38,"itemListElement":39},"BreadcrumbList",[40,44,48,51],{"item":41,"name":42,"@type":43,"position":20},"https://docshare.wps.com","Home","ListItem",{"item":45,"name":46,"@type":43,"position":47},"https://docshare.wps.com/document/","Document",2,{"item":49,"name":12,"@type":43,"position":50},"https://docshare.wps.com/document/research-report/",3,{"item":52,"name":13,"@type":43,"position":53},"https://docshare.wps.com/document/local-interpretation-of-machine-learning-models-in-remote-sensing-with-shap-the-case-of-global-climate-constraints-on-photosynthesis-phenology/127744/",4,{"url":52,"name":13,"@type":55,"author":56,"headline":13,"publisher":58,"fileFormat":61,"inLanguage":23,"description":14,"dateModified":62,"datePublished":63,"encodingFormat":61,"isAccessibleForFree":64,"interactionStatistic":65},"DigitalDocument",{"name":9,"@type":57},"Person",{"url":41,"name":59,"@type":60},"DocShare","Organization","application/pdf","2026-08-22","2026-08-05",true,{"@type":66,"interactionType":67,"userInteractionCount":20},"InteractionCounter",{"@type":68},"ViewAction",{"@type":70,"mainEntity":71},"FAQPage",[72,78,82],{"name":73,"@type":74,"acceptedAnswer":75},"Why is interpretability important for machine-learning remote sensing models?","Question",{"text":76,"@type":77},"Because these models often behave as “black boxes,” making the relationships between inputs and predicted outputs difficult to understand.","Answer",{"name":79,"@type":74,"acceptedAnswer":80},"How does SHAP contribute to interpreting the weather-SIF regression model?",{"text":81,"@type":77},"SHAP estimates the contribution of each input variable to individual predictions, enabling local interpretation of how inputs drive outcomes.",{"name":83,"@type":74,"acceptedAnswer":84},"Which climate variables were identified as limiting factors, and where?",{"text":85,"@type":77},"Radiation limited the photosynthetically active season in tropical rainforest, while vapor-pressure-deficit constrained start and end in tropical drylands; temperature limited the start in extra-tropical regions, and both temperature and radiation constrained the end.","https://schema.org",{"og:url":52,"og:type":88,"og:title":13,"og:site_name":59,"og:description":14},"article",{"robots":90,"canonical":52},"index,follow",{"doc_id":7,"site_id":24},{"code":4,"msg":5,"data":93},[94,98,102,106,110,115,120,123,128,131,135],{"id":20,"doc_module":4,"doc_module_name":46,"category_name":95,"show_sort_weight":96,"slug":97},"Story & Novel",90,"story-novel",{"id":47,"doc_module":4,"doc_module_name":46,"category_name":99,"show_sort_weight":100,"slug":101},"Literature",80,"literature",{"id":53,"doc_module":4,"doc_module_name":46,"category_name":103,"show_sort_weight":104,"slug":105},"Exam",70,"exam",{"id":107,"doc_module":4,"doc_module_name":46,"category_name":108,"show_sort_weight":29,"slug":109},5,"Comic","comic",{"id":111,"doc_module":4,"doc_module_name":46,"category_name":112,"show_sort_weight":113,"slug":114},6,"Technology",50,"technology",{"id":116,"doc_module":4,"doc_module_name":46,"category_name":117,"show_sort_weight":118,"slug":119},7,"Healthcare",40,"healthcare",{"id":11,"doc_module":4,"doc_module_name":46,"category_name":12,"show_sort_weight":121,"slug":122},30,"research-report",{"id":124,"doc_module":4,"doc_module_name":46,"category_name":125,"show_sort_weight":126,"slug":127},9,"Religion & Spirituality",20,"religion-spirituality",{"id":126,"doc_module":4,"doc_module_name":46,"category_name":129,"show_sort_weight":126,"slug":130},"World Cup","world-cup",{"id":132,"doc_module":4,"doc_module_name":46,"category_name":133,"show_sort_weight":132,"slug":134},10,"Lifestyle","lifestyle",{"id":136,"doc_module":4,"doc_module_name":46,"category_name":137,"show_sort_weight":107,"slug":138},19,"General","general"]