森林/非森林グローバル マップ(FNF)は、グローバル 25 m 分解能の PALSAR-2/PALSAR SAR モザイクの SAR 画像(後方散乱係数)を分類して生成されます。強い後方散乱ピクセルと弱い後方散乱ピクセルは、それぞれ「森林」と「非森林」に割り当てられます。ここで「森林」とは、面積が 0.5 ヘクタールを超え、森林被覆率が 10% を超える自然林と定義されます。この定義は、国連食糧農業機関(FAO)の定義と同じです。森林からのレーダー後方散乱は地域(気候帯)によって異なるため、森林/非森林の分類は、地域に依存する後方散乱のしきい値を使用して行われます。分類の精度は、現地写真と高解像度の光学衛星画像を使用して確認されます。詳しくは、プロバイダのデータセットの説明をご覧ください。
[[["わかりやすい","easyToUnderstand","thumb-up"],["問題の解決に役立った","solvedMyProblem","thumb-up"],["その他","otherUp","thumb-up"]],[["必要な情報がない","missingTheInformationINeed","thumb-down"],["複雑すぎる / 手順が多すぎる","tooComplicatedTooManySteps","thumb-down"],["最新ではない","outOfDate","thumb-down"],["翻訳に関する問題","translationIssue","thumb-down"],["サンプル / コードに問題がある","samplesCodeIssue","thumb-down"],["その他","otherDown","thumb-down"]],[],[[["\u003cp\u003eThe JAXA/ALOS/PALSAR/YEARLY/FNF dataset provides a global forest/non-forest map at 25-meter resolution, classifying areas as forest, non-forest, or water based on SAR backscatter data.\u003c/p\u003e\n"],["\u003cp\u003eThis dataset covers the period from 2007 to 2018 and utilizes a region-dependent threshold for classification to account for variations in radar backscatter across different climate zones.\u003c/p\u003e\n"],["\u003cp\u003eUsers should be aware that backscatter values might differ significantly in high-latitude forest areas due to seasonal changes, potentially affecting the classification accuracy.\u003c/p\u003e\n"],["\u003cp\u003eA newer version of the dataset, JAXA/ALOS/PALSAR/YEARLY/FNF4, offers an updated classification with four classes for the years 2017-2020.\u003c/p\u003e\n"],["\u003cp\u003eJAXA retains ownership of the dataset, and users are required to acknowledge the data source when publishing any results derived from it.\u003c/p\u003e\n"]]],["The JAXA EORC provides a global forest/non-forest (FNF) map dataset from 2007 to 2018. This dataset uses 25m resolution PALSAR-2/PALSAR SAR mosaic images, classifying pixels based on backscatter strength into \"forest,\" \"non-forest,\" and \"water.\" Forest is defined as natural forest with an area over 0.5 ha and cover exceeding 10%. The dataset is accessible via Google Earth Engine using the provided snippet and is classified with an accuracy determined through in-situ and high-resolution optical images.\n"],null,[]]