Why ecosystem characteristics predicted from remotely sensed data are unbiased and biased at the same time – and how this affects applications
Remotely sensed data are frequently used for predicting and mapping ecosystem characteristics, and spatially explicit wall-to-wall information is sometimes proposed as the best possible source of information for decision-making. However, wall-to-wall information typically relies on model-based predi...
| Main Authors: | , , , , , , , , , , , , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
KeAi Communications Co., Ltd.
2024-01-01
|
| Series: | Forest Ecosystems |
| Subjects: | |
| Online Access: | http://www.sciencedirect.com/science/article/pii/S2197562023000957 |
| _version_ | 1827049835349409792 |
|---|---|
| author | Göran Ståhl Terje Gobakken Svetlana Saarela Henrik J. Persson Magnus Ekström Sean P. Healey Zhiqiang Yang Johan Holmgren Eva Lindberg Kenneth Nyström Emanuele Papucci Patrik Ulvdal Hans Ole Ørka Erik Næsset Zhengyang Hou Håkan Olsson Ronald E. McRoberts |
| author_facet | Göran Ståhl Terje Gobakken Svetlana Saarela Henrik J. Persson Magnus Ekström Sean P. Healey Zhiqiang Yang Johan Holmgren Eva Lindberg Kenneth Nyström Emanuele Papucci Patrik Ulvdal Hans Ole Ørka Erik Næsset Zhengyang Hou Håkan Olsson Ronald E. McRoberts |
| author_sort | Göran Ståhl |
| collection | DOAJ |
| description | Remotely sensed data are frequently used for predicting and mapping ecosystem characteristics, and spatially explicit wall-to-wall information is sometimes proposed as the best possible source of information for decision-making. However, wall-to-wall information typically relies on model-based prediction, and several features of model-based prediction should be understood before extensively relying on this type of information. One such feature is that model-based predictors can be considered both unbiased and biased at the same time, which has important implications in several areas of application. In this discussion paper, we first describe the conventional model-unbiasedness paradigm that underpins most prediction techniques using remotely sensed (or other) auxiliary data. From this point of view, model-based predictors are typically unbiased. Secondly, we show that for specific domains, identified based on their true values, the same model-based predictors can be considered biased, and sometimes severely so.We suggest distinguishing between conventional model-bias, defined in the statistical literature as the difference between the expected value of a predictor and the expected value of the quantity being predicted, and design-bias of model-based estimators, defined as the difference between the expected value of a model-based estimator and the true value of the quantity being predicted. We show that model-based estimators (or predictors) are typically design-biased, and that there is a trend in the design-bias from overestimating small true values to underestimating large true values. Further, we give examples of applications where this is important to acknowledge and to potentially make adjustments to correct for the design-bias trend. We argue that relying entirely on conventional model-unbiasedness may lead to mistakes in several areas of application that use predictions from remotely sensed data. |
| first_indexed | 2024-03-08T14:21:28Z |
| format | Article |
| id | doaj.art-ac8f7042e23e48c4a21d47f6b234b47b |
| institution | Directory Open Access Journal |
| issn | 2197-5620 |
| language | English |
| last_indexed | 2025-02-18T15:50:42Z |
| publishDate | 2024-01-01 |
| publisher | KeAi Communications Co., Ltd. |
| record_format | Article |
| series | Forest Ecosystems |
| spelling | doaj.art-ac8f7042e23e48c4a21d47f6b234b47b2024-10-25T04:23:11ZengKeAi Communications Co., Ltd.Forest Ecosystems2197-56202024-01-0111100164Why ecosystem characteristics predicted from remotely sensed data are unbiased and biased at the same time – and how this affects applicationsGöran Ståhl0Terje Gobakken1Svetlana Saarela2Henrik J. Persson3Magnus Ekström4Sean P. Healey5Zhiqiang Yang6Johan Holmgren7Eva Lindberg8Kenneth Nyström9Emanuele Papucci10Patrik Ulvdal11Hans Ole Ørka12Erik Næsset13Zhengyang Hou14Håkan Olsson15Ronald E. McRoberts16Department of Forest Resource Management, Swedish University of Agricultural Sciences, 901 83, Umeå, Sweden; Corresponding author. Swedish University of Agricultural Sciences, 901 83, Umeå, Sweden.Faculty of Environmental Sciences and Natural Resource Management, Norwegian University of Life Sciences, Ås, NorwayFaculty of Environmental Sciences and Natural Resource Management, Norwegian University of Life Sciences, Ås, NorwayDepartment of Forest Resource Management, Swedish University of Agricultural Sciences, 901 83, Umeå, SwedenDepartment of Forest Resource Management, Swedish University of Agricultural Sciences, 901 83, Umeå, SwedenUSDA Forest Service, Rocky Mountain Research Station, Ogden, UT, USAUSDA Forest Service, Rocky Mountain Research Station, Ogden, UT, USADepartment of Forest Resource Management, Swedish University of Agricultural Sciences, 901 83, Umeå, SwedenDepartment of Forest Resource Management, Swedish University of Agricultural Sciences, 901 83, Umeå, SwedenDepartment of Forest Resource Management, Swedish University of Agricultural Sciences, 901 83, Umeå, SwedenDepartment of Forest Resource Management, Swedish University of Agricultural Sciences, 901 83, Umeå, SwedenDepartment of Forest Resource Management, Swedish University of Agricultural Sciences, 901 83, Umeå, SwedenFaculty of Environmental Sciences and Natural Resource Management, Norwegian University of Life Sciences, Ås, NorwayFaculty of Environmental Sciences and Natural Resource Management, Norwegian University of Life Sciences, Ås, NorwayThe Key Laboratory for Silviculture and Conservation of Ministry of Education, Beijing Forestry University, Beijing, 100083, ChinaDepartment of Forest Resource Management, Swedish University of Agricultural Sciences, 901 83, Umeå, SwedenDepartment of Forest Resources, University of Minnesota, St. Paul, MN, USARemotely sensed data are frequently used for predicting and mapping ecosystem characteristics, and spatially explicit wall-to-wall information is sometimes proposed as the best possible source of information for decision-making. However, wall-to-wall information typically relies on model-based prediction, and several features of model-based prediction should be understood before extensively relying on this type of information. One such feature is that model-based predictors can be considered both unbiased and biased at the same time, which has important implications in several areas of application. In this discussion paper, we first describe the conventional model-unbiasedness paradigm that underpins most prediction techniques using remotely sensed (or other) auxiliary data. From this point of view, model-based predictors are typically unbiased. Secondly, we show that for specific domains, identified based on their true values, the same model-based predictors can be considered biased, and sometimes severely so.We suggest distinguishing between conventional model-bias, defined in the statistical literature as the difference between the expected value of a predictor and the expected value of the quantity being predicted, and design-bias of model-based estimators, defined as the difference between the expected value of a model-based estimator and the true value of the quantity being predicted. We show that model-based estimators (or predictors) are typically design-biased, and that there is a trend in the design-bias from overestimating small true values to underestimating large true values. Further, we give examples of applications where this is important to acknowledge and to potentially make adjustments to correct for the design-bias trend. We argue that relying entirely on conventional model-unbiasedness may lead to mistakes in several areas of application that use predictions from remotely sensed data.http://www.sciencedirect.com/science/article/pii/S2197562023000957BiasModel-based inferenceDesign-based inference |
| spellingShingle | Göran Ståhl Terje Gobakken Svetlana Saarela Henrik J. Persson Magnus Ekström Sean P. Healey Zhiqiang Yang Johan Holmgren Eva Lindberg Kenneth Nyström Emanuele Papucci Patrik Ulvdal Hans Ole Ørka Erik Næsset Zhengyang Hou Håkan Olsson Ronald E. McRoberts Why ecosystem characteristics predicted from remotely sensed data are unbiased and biased at the same time – and how this affects applications Forest Ecosystems Bias Model-based inference Design-based inference |
| title | Why ecosystem characteristics predicted from remotely sensed data are unbiased and biased at the same time – and how this affects applications |
| title_full | Why ecosystem characteristics predicted from remotely sensed data are unbiased and biased at the same time – and how this affects applications |
| title_fullStr | Why ecosystem characteristics predicted from remotely sensed data are unbiased and biased at the same time – and how this affects applications |
| title_full_unstemmed | Why ecosystem characteristics predicted from remotely sensed data are unbiased and biased at the same time – and how this affects applications |
| title_short | Why ecosystem characteristics predicted from remotely sensed data are unbiased and biased at the same time – and how this affects applications |
| title_sort | why ecosystem characteristics predicted from remotely sensed data are unbiased and biased at the same time and how this affects applications |
| topic | Bias Model-based inference Design-based inference |
| url | http://www.sciencedirect.com/science/article/pii/S2197562023000957 |
| work_keys_str_mv | AT goranstahl whyecosystemcharacteristicspredictedfromremotelysenseddataareunbiasedandbiasedatthesametimeandhowthisaffectsapplications AT terjegobakken whyecosystemcharacteristicspredictedfromremotelysenseddataareunbiasedandbiasedatthesametimeandhowthisaffectsapplications AT svetlanasaarela whyecosystemcharacteristicspredictedfromremotelysenseddataareunbiasedandbiasedatthesametimeandhowthisaffectsapplications AT henrikjpersson whyecosystemcharacteristicspredictedfromremotelysenseddataareunbiasedandbiasedatthesametimeandhowthisaffectsapplications AT magnusekstrom whyecosystemcharacteristicspredictedfromremotelysenseddataareunbiasedandbiasedatthesametimeandhowthisaffectsapplications AT seanphealey whyecosystemcharacteristicspredictedfromremotelysenseddataareunbiasedandbiasedatthesametimeandhowthisaffectsapplications AT zhiqiangyang whyecosystemcharacteristicspredictedfromremotelysenseddataareunbiasedandbiasedatthesametimeandhowthisaffectsapplications AT johanholmgren whyecosystemcharacteristicspredictedfromremotelysenseddataareunbiasedandbiasedatthesametimeandhowthisaffectsapplications AT evalindberg whyecosystemcharacteristicspredictedfromremotelysenseddataareunbiasedandbiasedatthesametimeandhowthisaffectsapplications AT kennethnystrom whyecosystemcharacteristicspredictedfromremotelysenseddataareunbiasedandbiasedatthesametimeandhowthisaffectsapplications AT emanuelepapucci whyecosystemcharacteristicspredictedfromremotelysenseddataareunbiasedandbiasedatthesametimeandhowthisaffectsapplications AT patrikulvdal whyecosystemcharacteristicspredictedfromremotelysenseddataareunbiasedandbiasedatthesametimeandhowthisaffectsapplications AT hansoleørka whyecosystemcharacteristicspredictedfromremotelysenseddataareunbiasedandbiasedatthesametimeandhowthisaffectsapplications AT eriknæsset whyecosystemcharacteristicspredictedfromremotelysenseddataareunbiasedandbiasedatthesametimeandhowthisaffectsapplications AT zhengyanghou whyecosystemcharacteristicspredictedfromremotelysenseddataareunbiasedandbiasedatthesametimeandhowthisaffectsapplications AT hakanolsson whyecosystemcharacteristicspredictedfromremotelysenseddataareunbiasedandbiasedatthesametimeandhowthisaffectsapplications AT ronaldemcroberts whyecosystemcharacteristicspredictedfromremotelysenseddataareunbiasedandbiasedatthesametimeandhowthisaffectsapplications |