backend/src/prediction/predictionService.ts

@@ -442,8 +442,14 @@ interface TrajectoryTimeStep {
   hydr_grid?: TrajectoryHydrGrid;
 }
 
+// ALGO_CD → 프론트엔드 모델명 매핑
+const ALGO_CD_TO_MODEL: Record<string, string> = {
+  'OPENDRIFT': 'OpenDrift',
+  'POSEIDON':  'POSEIDON',
+};
+
 interface TrajectoryResult {
-  trajectory: Array<{ lat: number; lon: number; time: number; particle: number; stranded?: 0 | 1 }>;
+  trajectory: Array<{ lat: number; lon: number; time: number; particle: number; stranded?: 0 | 1; model: string }>;
   summary: {
     remainingVolume: number;
     weatheredVolume: number;
@@ -451,12 +457,12 @@ interface TrajectoryResult {
     beachedVolume: number;
     pollutionCoastLength: number;
   };
-  centerPoints: Array<{ lat: number; lon: number; time: number }>;
+  centerPoints: Array<{ lat: number; lon: number; time: number; model: string }>;
   windData: TrajectoryWindPoint[][];
   hydrData: ({ value: [number[][], number[][]]; grid: TrajectoryHydrGrid } | null)[];
 }
 
-function transformTrajectoryResult(rawResult: TrajectoryTimeStep[]): TrajectoryResult {
+function transformTrajectoryResult(rawResult: TrajectoryTimeStep[], model: string): TrajectoryResult {
   const trajectory = rawResult.flatMap((step, stepIdx) =>
     step.particles.map((p, i) => ({
       lat: p.lat,
@@ -464,6 +470,7 @@ function transformTrajectoryResult(rawResult: TrajectoryTimeStep[]): TrajectoryR
       time: stepIdx,
       particle: i,
       stranded: p.stranded,
+      model,
     }))
   );
   const lastStep = rawResult[rawResult.length - 1];
@@ -477,10 +484,10 @@ function transformTrajectoryResult(rawResult: TrajectoryTimeStep[]): TrajectoryR
   const centerPoints = rawResult
     .map((step, stepIdx) =>
       step.center_lat != null && step.center_lon != null
-        ? { lat: step.center_lat, lon: step.center_lon, time: stepIdx }
+        ? { lat: step.center_lat, lon: step.center_lon, time: stepIdx, model }
         : null
     )
-    .filter((p): p is { lat: number; lon: number; time: number } => p !== null);
+    .filter((p): p is { lat: number; lon: number; time: number; model: string } => p !== null);
   const windData = rawResult.map((step) => step.wind_data ?? []);
   const hydrData = rawResult.map((step) =>
     step.hydr_data && step.hydr_grid
@@ -491,14 +498,52 @@ function transformTrajectoryResult(rawResult: TrajectoryTimeStep[]): TrajectoryR
 }
 
 export async function getAnalysisTrajectory(acdntSn: number): Promise<TrajectoryResult | null> {
+  // 완료된 모든 모델(OPENDRIFT, POSEIDON) 결과 조회
   const sql = `
-    SELECT RSLT_DATA FROM wing.PRED_EXEC
-    WHERE ACDNT_SN = $1 AND ALGO_CD = 'OPENDRIFT' AND EXEC_STTS_CD = 'COMPLETED'
-    ORDER BY CMPL_DTM DESC LIMIT 1
+    SELECT ALGO_CD, RSLT_DATA, CMPL_DTM FROM wing.PRED_EXEC
+    WHERE ACDNT_SN = $1
+      AND ALGO_CD IN ('OPENDRIFT', 'POSEIDON')
+      AND EXEC_STTS_CD = 'COMPLETED'
+    ORDER BY CMPL_DTM DESC
   `;
   const { rows } = await wingPool.query(sql, [acdntSn]);
-  if (rows.length === 0 || !rows[0].rslt_data) return null;
-  return transformTrajectoryResult(rows[0].rslt_data as TrajectoryTimeStep[]);
+  if (rows.length === 0) return null;
+
+  // 모든 모델의 파티클을 하나의 배열로 병합
+  let mergedTrajectory: TrajectoryResult['trajectory'] = [];
+  let allCenterPoints: TrajectoryResult['centerPoints'] = [];
+
+  // summary/windData/hydrData: 가장 최근 완료된 OpenDrift 기준, 없으면 POSEIDON 기준
+  let baseResult: TrajectoryResult | null = null;
+
+  // OpenDrift 우선, 없으면 POSEIDON 선택 (ORDER BY CMPL_DTM DESC이므로 첫 번째 행이 가장 최근)
+  const opendriftRow = (rows as Array<Record<string, unknown>>).find((r) => r['algo_cd'] === 'OPENDRIFT');
+  const poseidonRow = (rows as Array<Record<string, unknown>>).find((r) => r['algo_cd'] === 'POSEIDON');
+  const baseRow = opendriftRow ?? poseidonRow ?? null;
+
+  for (const row of rows as Array<Record<string, unknown>>) {
+    if (!row['rslt_data']) continue;
+    const algoCd = String(row['algo_cd'] ?? '');
+    const modelName = ALGO_CD_TO_MODEL[algoCd] ?? algoCd;
+    const parsed = transformTrajectoryResult(row['rslt_data'] as TrajectoryTimeStep[], modelName);
+    mergedTrajectory = mergedTrajectory.concat(parsed.trajectory);
+    allCenterPoints = allCenterPoints.concat(parsed.centerPoints);
+
+    // 기준 행의 결과를 baseResult로 사용
+    if (row === baseRow) {
+      baseResult = parsed;
+    }
+  }
+
+  if (!baseResult) return null;
+
+  return {
+    trajectory: mergedTrajectory,
+    summary: baseResult.summary,
+    centerPoints: allCenterPoints,
+    windData: baseResult.windData,
+    hydrData: baseResult.hydrData,
+  };
 }
 
 export async function listBoomLines(acdntSn: number): Promise<BoomLineItem[]> {

backend/src/routes/simulation.ts

@@ -11,6 +11,7 @@ import {
 const router = Router()
 
 const PYTHON_API_URL = process.env.PYTHON_API_URL ?? 'http://localhost:5003'
+const POSEIDON_API_URL = process.env.POSEIDON_API_URL ?? 'http://localhost:5004'
 const POLL_INTERVAL_MS = 3000
 const POLL_TIMEOUT_MS = 30 * 60 * 1000 // 30분
 
@@ -71,20 +72,38 @@ async function rollbackNewRecords(
   }
 }
 
+// 모델명 → ALGO_CD 매핑
+const MODEL_ALGO_CD_MAP: Record<string, string> = {
+  'OpenDrift': 'OPENDRIFT',
+  'POSEIDON':  'POSEIDON',
+}
+
+// 모델명 → API URL 매핑
+const MODEL_API_URL_MAP: Record<string, string> = {
+  'OpenDrift': PYTHON_API_URL,
+  'POSEIDON':  POSEIDON_API_URL,
+}
+
 // ============================================================
 // POST /api/simulation/run
-// 확산 시뮬레이션 실행 (OpenDrift)
+// 확산 시뮬레이션 실행 (다중 모델 지원: OpenDrift, POSEIDON)
 // ============================================================
 /**
- * OpenDrift 확산 시뮬레이션을 실행한다.
- * Python FastAPI 서버에 작업을 제출하고 job_id를 받아
- * 백그라운드에서 폴링하며 결과를 DB에 저장한다.
- * 프론트엔드는 execSn으로 GET /status/:execSn을 폴링하여 결과를 수신한다.
+ * 선택된 모델(OpenDrift, POSEIDON)로 확산 시뮬레이션을 실행한다.
+ * 각 모델에 대해 PRED_EXEC 레코드를 별도 생성하고 Python API에 병렬 제출한다.
+ * KOSPS 모델은 PRED_EXEC INSERT(PENDING)만 수행하고 외부 API 연동은 하지 않는다.
+ * 프론트엔드는 execSns 배열의 각 execSn으로 GET /status/:execSn을 폴링하여 결과를 수신한다.
  */
 router.post('/run', requireAuth, async (req: Request, res: Response) => {
   try {
     const { acdntSn: rawAcdntSn, acdntNm, spillUnit, spillTypeCd,
-            lat, lon, runTime, matTy, matVol, spillTime, startTime } = req.body
+            lat, lon, runTime, matTy, matVol, spillTime, startTime,
+            models: rawModels } = req.body
+
+    // 실행할 모델 목록 (기본값: OpenDrift)
+    const requestedModels: string[] = Array.isArray(rawModels) && rawModels.length > 0
+      ? (rawModels as string[])
+      : ['OpenDrift']
 
     // 1. 필수 파라미터 검증
     if (lat === undefined || lon === undefined || runTime === undefined) {
@@ -117,21 +136,24 @@ router.post('/run', requireAuth, async (req: Request, res: Response) => {
     }
 
     // 2. Python NC 파일 존재 여부 확인 (ACDNT 생성 전에 수행하여 고아 레코드 방지)
-    try {
-      const checkRes = await fetch(`${PYTHON_API_URL}/check-nc`, {
-        method: 'POST',
-        headers: { 'Content-Type': 'application/json' },
-        body: JSON.stringify({ lat, lon, startTime }),
-        signal: AbortSignal.timeout(5000),
-      })
-      if (!checkRes.ok) {
-        return res.status(409).json({
-          error: '해당 좌표의 해양 기상 데이터가 없습니다.',
-          message: 'NC 파일이 준비되지 않았습니다.',
+    // OpenDrift 모델이 포함된 경우에만 check-nc 수행
+    if (requestedModels.includes('OpenDrift')) {
+      try {
+        const checkRes = await fetch(`${PYTHON_API_URL}/check-nc`, {
+          method: 'POST',
+          headers: { 'Content-Type': 'application/json' },
+          body: JSON.stringify({ lat, lon, startTime }),
+          signal: AbortSignal.timeout(5000),
         })
+        if (!checkRes.ok) {
+          return res.status(409).json({
+            error: '해당 좌표의 해양 기상 데이터가 없습니다.',
+            message: 'NC 파일이 준비되지 않았습니다.',
+          })
+        }
+      } catch {
+        // Python 서버 미기동 — 5번에서 처리
       }
-    } catch {
-      // Python 서버 미기동 — 5번에서 처리
     }
 
     // 1-B. acdntSn 미제공 시 ACDNT + SPIL_DATA 생성
@@ -199,83 +221,131 @@ router.post('/run', requireAuth, async (req: Request, res: Response) => {
       }
     }
 
-    // 4. PRED_EXEC INSERT (PENDING) — ACDNT_SN 포함 (NOT NULL FK)
-    const execNm = `EXPC_${Date.now()}`
-    let predExecSn: number
-    try {
-      const insertRes = await wingPool.query(
-        `INSERT INTO wing.PRED_EXEC (ACDNT_SN, SPIL_DATA_SN, ALGO_CD, EXEC_STTS_CD, EXEC_NM, BGNG_DTM)
-         VALUES ($1, $2, 'OPENDRIFT', 'PENDING', $3, NOW())
-         RETURNING PRED_EXEC_SN`,
-        [resolvedAcdntSn, resolvedSpilDataSn, execNm]
-      )
-      predExecSn = insertRes.rows[0].pred_exec_sn as number
-    } catch (dbErr) {
-      console.error('[simulation] PRED_EXEC INSERT 실패:', dbErr)
-      return res.status(500).json({ error: '분석 기록 생성 실패' })
-    }
-
     // matTy 변환: 한국어 유종 → OpenDrift 유종 코드
     // 매핑 대상이 아니면 원본 값 그대로 사용 (영문 직접 입력 대응)
     const odMatTy = matTy !== undefined ? (OIL_TYPE_MAP[matTy as string] ?? (matTy as string)) : undefined
 
-    // 5. Python /run-model 호출
-    let jobId: string
-    try {
-      const pythonRes = await fetch(`${PYTHON_API_URL}/run-model`, {
-        method: 'POST',
-        headers: { 'Content-Type': 'application/json' },
-        body: JSON.stringify({
-          lat,
-          lon,
-          startTime,
-          runTime,
-          matTy: odMatTy,
-          matVol,
-          spillTime,
-          name: execNm,
-        }),
-        signal: AbortSignal.timeout(10000),
-      })
+    // 4. 각 모델별 PRED_EXEC INSERT 및 API 호출 (병렬)
+    // KOSPS: PRED_EXEC PENDING 생성만 하고 배열에서 제외 (외부 API 미연동)
+    const execNmBase = `EXPC_${Date.now()}`
+    const execSns: Array<{ model: string; execSn: number }> = []
 
-      if (pythonRes.status === 503) {
-        const errData = await pythonRes.json() as { error?: string }
-        await wingPool.query(
-          `UPDATE wing.PRED_EXEC SET EXEC_STTS_CD='FAILED', ERR_MSG=$1, CMPL_DTM=NOW() WHERE PRED_EXEC_SN=$2`,
-          [errData.error || '분석 서버 포화', predExecSn]
+    // KOSPS 처리: PRED_EXEC INSERT(PENDING)만 수행
+    if (requestedModels.includes('KOSPS')) {
+      try {
+        const kospsExecNm = `${execNmBase}_KOSPS`
+        const insertRes = await wingPool.query(
+          `INSERT INTO wing.PRED_EXEC (ACDNT_SN, SPIL_DATA_SN, ALGO_CD, EXEC_STTS_CD, EXEC_NM, BGNG_DTM)
+           VALUES ($1, $2, 'KOSPS', 'PENDING', $3, NOW())
+           RETURNING PRED_EXEC_SN`,
+          [resolvedAcdntSn, resolvedSpilDataSn, kospsExecNm]
         )
-        await rollbackNewRecords(predExecSn, newlyCreatedSpilDataSn, newlyCreatedAcdntSn)
-        return res.status(503).json({ error: errData.error || '분석 서버가 사용 중입니다. 잠시 후 재시도해 주세요.' })
+        execSns.push({ model: 'KOSPS', execSn: insertRes.rows[0].pred_exec_sn as number })
+      } catch (dbErr) {
+        console.error('[simulation] KOSPS PRED_EXEC INSERT 실패:', dbErr)
       }
-
-      if (!pythonRes.ok) {
-        throw new Error(`Python 서버 응답 오류: ${pythonRes.status}`)
-      }
-
-      const pythonData = await pythonRes.json() as { job_id: string }
-      jobId = pythonData.job_id
-    } catch {
-      await wingPool.query(
-        `UPDATE wing.PRED_EXEC SET EXEC_STTS_CD='FAILED', ERR_MSG='Python 분석 서버에 연결할 수 없습니다.', CMPL_DTM=NOW() WHERE PRED_EXEC_SN=$1`,
-        [predExecSn]
-      )
-      await rollbackNewRecords(predExecSn, newlyCreatedSpilDataSn, newlyCreatedAcdntSn)
-      return res.status(503).json({ error: 'Python 분석 서버에 연결할 수 없습니다.' })
     }
 
-    // 6. RUNNING 업데이트
-    await wingPool.query(
-      `UPDATE wing.PRED_EXEC SET EXEC_STTS_CD='RUNNING' WHERE PRED_EXEC_SN=$1`,
-      [predExecSn]
+    // API 연동 모델 필터링 (KOSPS 제외)
+    const apiModels = requestedModels.filter((m) => m !== 'KOSPS' && MODEL_ALGO_CD_MAP[m] !== undefined)
+
+    // 각 모델에 대해 PRED_EXEC INSERT → /run-model 호출
+    await Promise.all(
+      apiModels.map(async (model) => {
+        const algoCd = MODEL_ALGO_CD_MAP[model]
+        const apiUrl = MODEL_API_URL_MAP[model]
+        const execNm = `${execNmBase}_${algoCd}`
+
+        // PRED_EXEC INSERT (PENDING)
+        let predExecSn: number
+        try {
+          const insertRes = await wingPool.query(
+            `INSERT INTO wing.PRED_EXEC (ACDNT_SN, SPIL_DATA_SN, ALGO_CD, EXEC_STTS_CD, EXEC_NM, BGNG_DTM)
+             VALUES ($1, $2, $3, 'PENDING', $4, NOW())
+             RETURNING PRED_EXEC_SN`,
+            [resolvedAcdntSn, resolvedSpilDataSn, algoCd, execNm]
+          )
+          predExecSn = insertRes.rows[0].pred_exec_sn as number
+        } catch (dbErr) {
+          console.error(`[simulation] ${model} PRED_EXEC INSERT 실패:`, dbErr)
+          return
+        }
+
+        execSns.push({ model, execSn: predExecSn })
+
+        // Python /run-model 호출
+        let jobId: string
+        try {
+          const pythonRes = await fetch(`${apiUrl}/run-model`, {
+            method: 'POST',
+            headers: { 'Content-Type': 'application/json' },
+            body: JSON.stringify({
+              lat,
+              lon,
+              startTime,
+              runTime,
+              matTy: odMatTy,
+              matVol,
+              spillTime,
+              name: execNm,
+            }),
+            signal: AbortSignal.timeout(10000),
+          })
+
+          if (pythonRes.status === 503) {
+            const errData = await pythonRes.json() as { error?: string }
+            await wingPool.query(
+              `UPDATE wing.PRED_EXEC SET EXEC_STTS_CD='FAILED', ERR_MSG=$1, CMPL_DTM=NOW() WHERE PRED_EXEC_SN=$2`,
+              [errData.error || '분석 서버 포화', predExecSn]
+            )
+            await rollbackNewRecords(predExecSn, newlyCreatedSpilDataSn, newlyCreatedAcdntSn)
+            return
+          }
+
+          if (!pythonRes.ok) {
+            throw new Error(`Python 서버 응답 오류: ${pythonRes.status}`)
+          }
+
+          const pythonData = await pythonRes.json() as { job_id: string }
+          jobId = pythonData.job_id
+        } catch {
+          await wingPool.query(
+            `UPDATE wing.PRED_EXEC SET EXEC_STTS_CD='FAILED', ERR_MSG='Python 분석 서버에 연결할 수 없습니다.', CMPL_DTM=NOW() WHERE PRED_EXEC_SN=$1`,
+            [predExecSn]
+          )
+          // 이 모델의 PRED_EXEC만 롤백 (다른 모델은 계속 진행)
+          await rollbackNewRecords(predExecSn, null, null)
+          return
+        }
+
+        // RUNNING 업데이트
+        await wingPool.query(
+          `UPDATE wing.PRED_EXEC SET EXEC_STTS_CD='RUNNING' WHERE PRED_EXEC_SN=$1`,
+          [predExecSn]
+        )
+
+        // 백그라운드 폴링 시작
+        pollAndSaveModel(jobId, predExecSn, apiUrl, algoCd).catch((err: unknown) =>
+          console.error(`[simulation] ${model} pollAndSaveModel 오류:`, err)
+        )
+      })
     )
 
-    // 7. 즉시 응답 (프론트엔드는 execSn으로 폴링, acdntSn은 신규 생성 사고 추적용)
-    res.json({ success: true, execSn: predExecSn, acdntSn: resolvedAcdntSn, status: 'RUNNING' })
+    // ACDNT/SPIL_DATA가 신규 생성됐으나 모든 모델이 실패한 경우 롤백
+    const hasRunning = execSns.some(({ model }) => model !== 'KOSPS')
+    if (!hasRunning && newlyCreatedAcdntSn !== null) {
+      await rollbackNewRecords(null, newlyCreatedSpilDataSn, newlyCreatedAcdntSn)
+      return res.status(503).json({ error: '분석 서버에 연결할 수 없습니다.' })
+    }
 
-    // 8. 백그라운드 폴링 시작
-    pollAndSave(jobId, predExecSn).catch((err: unknown) =>
-      console.error('[simulation] pollAndSave 오류:', err)
-    )
+    // 즉시 응답 (하위 호환을 위해 execSn도 포함)
+    res.json({
+      success: true,
+      execSns,
+      execSn: execSns[0]?.execSn ?? 0,
+      acdntSn: resolvedAcdntSn,
+      status: 'RUNNING',
+    })
   } catch {
     res.status(500).json({ error: '시뮬레이션 실행 실패', message: '서버 내부 오류가 발생했습니다.' })
   }
@@ -297,7 +367,7 @@ router.get('/status/:execSn', requireAuth, async (req: Request, res: Response) =
 
   try {
     const result = await wingPool.query(
-      `SELECT pe.EXEC_STTS_CD, pe.RSLT_DATA, pe.ERR_MSG, pe.BGNG_DTM, sd.FCST_HR,
+      `SELECT pe.EXEC_STTS_CD, pe.RSLT_DATA, pe.ERR_MSG, pe.BGNG_DTM, pe.ALGO_CD, sd.FCST_HR,
               (
                 SELECT AVG(hist.REQD_SEC::FLOAT / hsd.FCST_HR)
                 FROM wing.PRED_EXEC hist
@@ -328,7 +398,9 @@ router.get('/status/:execSn', requireAuth, async (req: Request, res: Response) =
     const status = statusMap[dbStatus] ?? dbStatus
 
     if (status === 'DONE' && row.rslt_data) {
-      const { trajectory, summary, centerPoints, windData, hydrData } = transformResult(row.rslt_data as PythonTimeStep[])
+      const algoCd = String(row.algo_cd ?? '')
+      const modelName = ALGO_CD_TO_MODEL_NAME[algoCd] ?? algoCd
+      const { trajectory, summary, centerPoints, windData, hydrData } = transformResult(row.rslt_data as PythonTimeStep[], modelName)
       return res.json({ status, trajectory, summary, centerPoints, windData, hydrData })
     }
 
@@ -356,14 +428,14 @@ router.get('/status/:execSn', requireAuth, async (req: Request, res: Response) =
 // ============================================================
 // 백그라운드 폴링
 // ============================================================
-async function pollAndSave(jobId: string, execSn: number): Promise<void> {
+async function pollAndSaveModel(jobId: string, execSn: number, apiUrl: string, algoCode: string): Promise<void> {
   const deadline = Date.now() + POLL_TIMEOUT_MS
 
   while (Date.now() < deadline) {
     await new Promise<void>(resolve => setTimeout(resolve, POLL_INTERVAL_MS))
 
     try {
-      const pollRes = await fetch(`${PYTHON_API_URL}/status/${jobId}`, {
+      const pollRes = await fetch(`${apiUrl}/status/${jobId}`, {
         signal: AbortSignal.timeout(5000),
       })
       if (!pollRes.ok) continue
@@ -446,7 +518,13 @@ interface PythonStatusResponse {
   error?: string
 }
 
-function transformResult(rawResult: PythonTimeStep[]) {
+// ALGO_CD → 프론트엔드 모델명 매핑
+const ALGO_CD_TO_MODEL_NAME: Record<string, string> = {
+  'OPENDRIFT': 'OpenDrift',
+  'POSEIDON':  'POSEIDON',
+}
+
+function transformResult(rawResult: PythonTimeStep[], model: string) {
   const trajectory = rawResult.flatMap((step, stepIdx) =>
     step.particles.map((p, i) => ({
       lat: p.lat,
@@ -467,10 +545,10 @@ function transformResult(rawResult: PythonTimeStep[]) {
   const centerPoints = rawResult
     .map((step, stepIdx) =>
       step.center_lat != null && step.center_lon != null
-        ? { lat: step.center_lat, lon: step.center_lon, time: stepIdx }
+        ? { lat: step.center_lat, lon: step.center_lon, time: stepIdx, model }
         : null
     )
-    .filter((p): p is { lat: number; lon: number; time: number } => p !== null)
+    .filter((p): p is { lat: number; lon: number; time: number; model: string } => p !== null)
   const windData = rawResult.map((step) => step.wind_data ?? [])
   const hydrData = rawResult.map((step) =>
     step.hydr_data && step.hydr_grid

docs/RELEASE-NOTES.md

@@ -4,6 +4,11 @@
 
 ## [Unreleased]
 
+## [2026-03-17]
+
+### 추가
+- 다중 모델 시뮬레이션 지원 (OpenDrift + POSEIDON 병렬 실행 및 결과 병합)
+
 ## [2026-03-16]
 
 ### 추가

frontend/src/common/components/layout/TopBar.tsx

@@ -78,7 +78,7 @@ export function TopBar({ activeTab, onTabChange }: TopBarProps) {
 
           {/* 실시간 상황관리 */}
           <button
-            onClick={() => window.open(import.meta.env.VITE_SITUATIONAL_URL ?? 'http://localhost:5174', '_blank')}
+            onClick={() => window.open(import.meta.env.VITE_SITUATIONAL_URL ?? 'https://kcg.gc-si.dev', '_blank')}
             className={`
               px-2.5 xl:px-4 py-2 rounded-sm text-[13px] transition-all duration-200
               font-korean tracking-[0.2px] font-semibold

frontend/src/common/components/map/MapView.tsx

@@ -321,7 +321,7 @@ interface MapViewProps {
   incidentCoord?: { lon: number; lat: number }
   isSelectingLocation?: boolean
   onMapClick?: (lon: number, lat: number) => void
-  oilTrajectory?: Array<{ lat: number; lon: number; time: number; particle?: number; model?: PredictionModel; stranded?: 0 | 1 }>
+  oilTrajectory?: Array<{ lat: number; lon: number; time: number; particle?: number; model?: string; stranded?: 0 | 1 }>
   selectedModels?: Set<PredictionModel>
   dispersionResult?: DispersionResult | null
   dispersionHeatmap?: Array<{ lon: number; lat: number; concentration: number }>
@@ -341,7 +341,7 @@ interface MapViewProps {
   sensitiveResources?: SensitiveResource[]
   flyToTarget?: { lng: number; lat: number; zoom?: number } | null
   fitBoundsTarget?: { north: number; south: number; east: number; west: number } | null
-  centerPoints?: Array<{ lat: number; lon: number; time: number }>
+  centerPoints?: Array<{ lat: number; lon: number; time: number; model?: string }>
   windData?: Array<Array<{ lat: number; lon: number; wind_speed: number; wind_direction: number }>>
   hydrData?: (HydrDataStep | null)[]
   // 외부 플레이어 제어 (prediction 하단 바에서 제어할 때 사용)
@@ -414,14 +414,15 @@ function MapPitchController({ threeD }: { threeD: boolean }) {
 }
 
 // 사고 지점 변경 시 지도 이동 (Map 내부 컴포넌트)
-function MapFlyToIncident({ lon, lat, onFlyEnd }: { lon?: number; lat?: number; onFlyEnd?: () => void }) {
+function MapFlyToIncident({ coord, onFlyEnd }: { coord?: { lon: number; lat: number }; onFlyEnd?: () => void }) {
   const { current: map } = useMap()
   const onFlyEndRef = useRef(onFlyEnd)
   useEffect(() => { onFlyEndRef.current = onFlyEnd }, [onFlyEnd])
 
   useEffect(() => {
-    if (!map || lon == null || lat == null) return
+    if (!map || !coord) return
 
+    const { lon, lat } = coord
     const doFly = () => {
       map.flyTo({ center: [lon, lat], zoom: 11, duration: 1200 })
       map.once('moveend', () => onFlyEndRef.current?.())
@@ -432,7 +433,7 @@ function MapFlyToIncident({ lon, lat, onFlyEnd }: { lon?: number; lat?: number;
     } else {
       map.once('load', doFly)
     }
-  }, [lon, lat, map])
+  }, [coord, map])  // 객체 참조 추적: 같은 좌표라도 새 객체면 effect 재실행
 
   return null
 }
@@ -596,15 +597,13 @@ export function MapView({
           getPosition: (d: (typeof visibleParticles)[0]) => [d.lon, d.lat],
           getRadius: 3,
           getFillColor: (d: (typeof visibleParticles)[0]) => {
-            // 1순위: stranded 입자 → showBeached=true 시 빨간색, false 시 회색
+            const modelKey = (d.model || Array.from(selectedModels)[0] || 'OpenDrift') as PredictionModel
+            // 1순위: stranded 입자 → showBeached=true 시 모델 색, false 시 회색
             if (d.stranded === 1) return showBeached
-              ? [239, 68, 68, 220] as [number, number, number, number]
+              ? hexToRgba(MODEL_COLORS[modelKey] || '#3b82f6', 180)
               : [130, 130, 130, 70] as [number, number, number, number]
             // 2순위: 현재 활성 스텝 → 모델 기본 색상
-            if (d.time === activeStep) {
-              const modelKey = d.model || Array.from(selectedModels)[0] || 'OpenDrift'
-              return hexToRgba(MODEL_COLORS[modelKey] || '#3b82f6', 180)
-            }
+            if (d.time === activeStep) return hexToRgba(MODEL_COLORS[modelKey] || '#3b82f6', 180)
             // 3순위: 과거 스텝 → 회색 + 투명
             return [130, 130, 130, 70] as [number, number, number, number]
           },
@@ -638,6 +637,31 @@ export function MapView({
       )
     }
 
+    // --- 육지부착 hollow ring (stranded 모양 구분) ---
+    const strandedParticles = showBeached ? visibleParticles.filter(p => p.stranded === 1) : []
+    if (strandedParticles.length > 0) {
+      result.push(
+        new ScatterplotLayer({
+          id: 'oil-stranded-ring',
+          data: strandedParticles,
+          getPosition: (d: (typeof strandedParticles)[0]) => [d.lon, d.lat],
+          stroked: true,
+          filled: false,
+          getLineColor: (d: (typeof strandedParticles)[0]) => {
+            const modelKey = (d.model || Array.from(selectedModels)[0] || 'OpenDrift') as PredictionModel
+            return hexToRgba(MODEL_COLORS[modelKey] || '#3b82f6', 255)
+          },
+          lineWidthMinPixels: 2,
+          getRadius: 4,
+          radiusMinPixels: 5,
+          radiusMaxPixels: 8,
+          updateTriggers: {
+            getLineColor: [selectedModels],
+          },
+        })
+      )
+    }
+
     // --- 오일펜스 라인 (PathLayer) ---
     if (boomLines.length > 0) {
       result.push(
@@ -1042,65 +1066,74 @@ export function MapView({
       )
     }
 
-    // --- 입자 중심점 이동 경로 (PathLayer + ScatterplotLayer) ---
+    // --- 입자 중심점 이동 경로 (모델별 PathLayer + ScatterplotLayer) ---
     const visibleCenters = centerPoints.filter(p => p.time <= currentTime)
-    if (visibleCenters.length >= 2) {
-      result.push(
-        new PathLayer({
-          id: 'center-path',
-          data: [{ path: visibleCenters.map(p => [p.lon, p.lat] as [number, number]) }],
-          getPath: (d: { path: [number, number][] }) => d.path,
-          getColor: (lightMode ? [0, 60, 150, 230] : [255, 220, 50, 200]) as [number, number, number, number],
-          getWidth: 2,
-          widthMinPixels: 2,
-          widthMaxPixels: 4,
-        })
-      )
-    }
     if (visibleCenters.length > 0) {
-      result.push(
-        new ScatterplotLayer({
-          id: 'center-points',
-          data: visibleCenters,
-          getPosition: (d: (typeof visibleCenters)[0]) => [d.lon, d.lat],
-          getRadius: 5,
-          getFillColor: (lightMode ? [0, 60, 150, 230] : [255, 220, 50, 230]) as [number, number, number, number],
-          radiusMinPixels: 4,
-          radiusMaxPixels: 8,
-          pickable: false,
-        })
-      )
-    }
+      // 모델별 그룹핑 (Record 사용 — Map 컴포넌트와 이름 충돌 회피)
+      const modelGroups: Record<string, typeof visibleCenters> = {}
+      visibleCenters.forEach(p => {
+        const key = p.model || 'OpenDrift'
+        if (!modelGroups[key]) modelGroups[key] = []
+        modelGroups[key].push(p)
+      })
 
-    // --- 시간 표시 라벨 (TextLayer) ---
-    if (visibleCenters.length > 0 && showTimeLabel) {
-      const baseTime = simulationStartTime ? new Date(simulationStartTime) : null;
-      const pad = (n: number) => String(n).padStart(2, '0');
-      result.push(
-        new TextLayer({
-          id: 'time-labels',
-          data: visibleCenters,
-          getPosition: (d: (typeof visibleCenters)[0]) => [d.lon, d.lat],
-          getText: (d: (typeof visibleCenters)[0]) => {
-            if (baseTime) {
-              const dt = new Date(baseTime.getTime() + d.time * 3600 * 1000);
-              return `${dt.getFullYear()}-${pad(dt.getMonth() + 1)}-${pad(dt.getDate())} ${pad(dt.getHours())}:${pad(dt.getMinutes())}`;
-            }
-            return `+${d.time}h`;
-          },
-          getSize: 12,
-          getColor: (lightMode ? [20, 40, 100, 240] : [255, 220, 50, 220]) as [number, number, number, number],
-          getPixelOffset: [0, 16] as [number, number],
-          fontWeight: 'bold',
-          outlineWidth: 2,
-          outlineColor: (lightMode ? [255, 255, 255, 180] : [15, 21, 36, 200]) as [number, number, number, number],
-          billboard: true,
-          sizeUnits: 'pixels' as const,
-          updateTriggers: {
-            getText: [simulationStartTime, currentTime],
-          },
-        })
-      )
+      Object.entries(modelGroups).forEach(([model, points]) => {
+        const modelColor = hexToRgba(MODEL_COLORS[model as PredictionModel] || '#06b6d4', 210)
+        if (points.length >= 2) {
+          result.push(
+            new PathLayer({
+              id: `center-path-${model}`,
+              data: [{ path: points.map((p: { lon: number; lat: number }) => [p.lon, p.lat] as [number, number]) }],
+              getPath: (d: { path: [number, number][] }) => d.path,
+              getColor: modelColor,
+              getWidth: 2,
+              widthMinPixels: 2,
+              widthMaxPixels: 4,
+            })
+          )
+        }
+        result.push(
+          new ScatterplotLayer({
+            id: `center-points-${model}`,
+            data: points,
+            getPosition: (d: { lon: number; lat: number }) => [d.lon, d.lat],
+            getRadius: 5,
+            getFillColor: modelColor,
+            radiusMinPixels: 4,
+            radiusMaxPixels: 8,
+            pickable: false,
+          })
+        )
+        if (showTimeLabel) {
+          const baseTime = simulationStartTime ? new Date(simulationStartTime) : null;
+          const pad = (n: number) => String(n).padStart(2, '0');
+          result.push(
+            new TextLayer({
+              id: `time-labels-${model}`,
+              data: points,
+              getPosition: (d: { lon: number; lat: number }) => [d.lon, d.lat],
+              getText: (d: { time: number }) => {
+                if (baseTime) {
+                  const dt = new Date(baseTime.getTime() + d.time * 3600 * 1000);
+                  return `${dt.getFullYear()}-${pad(dt.getMonth() + 1)}-${pad(dt.getDate())} ${pad(dt.getHours())}:${pad(dt.getMinutes())}`;
+                }
+                return `+${d.time}h`;
+              },
+              getSize: 12,
+              getColor: hexToRgba(MODEL_COLORS[model as PredictionModel] || '#06b6d4', 240),
+              getPixelOffset: [0, 16] as [number, number],
+              fontWeight: 'bold',
+              outlineWidth: 2,
+              outlineColor: (lightMode ? [255, 255, 255, 180] : [15, 21, 36, 200]) as [number, number, number, number],
+              billboard: true,
+              sizeUnits: 'pixels' as const,
+              updateTriggers: {
+                getText: [simulationStartTime, currentTime],
+              },
+            })
+          )
+        }
+      })
     }
 
     // --- 바람 화살표 (TextLayer) ---
@@ -1177,7 +1210,7 @@ export function MapView({
         {/* 3D 모드 pitch 제어 */}
         <MapPitchController threeD={mapToggles.threeD} />
         {/* 사고 지점 변경 시 지도 이동 */}
-        <MapFlyToIncident lon={flyToIncident?.lon} lat={flyToIncident?.lat} onFlyEnd={onIncidentFlyEnd} />
+        <MapFlyToIncident coord={flyToIncident} onFlyEnd={onIncidentFlyEnd} />
         {/* 외부에서 flyTo 트리거 */}
         <FlyToController flyToTarget={flyToTarget} />
         {/* 예측 완료 시 궤적 전체 범위로 fitBounds */}

frontend/src/tabs/aerial/components/OilAreaAnalysis.tsx

@@ -8,15 +8,59 @@ import { decimalToDMS } from '@common/utils/coordinates';
 
 const MAX_IMAGES = 6;
 
-interface GpsInfo {
+interface ImageExif {
   lat: number | null;
   lon: number | null;
+  altitude: number | null;
+  make: string | null;
+  model: string | null;
+  dateTime: Date | string | null;
+  exposureTime: number | null;
+  fNumber: number | null;
+  iso: number | null;
+  focalLength: number | null;
+  imageWidth: number | null;
+  imageHeight: number | null;
+}
+
+function formatFileSize(bytes?: number): string | null {
+  if (bytes == null) return null;
+  if (bytes < 1024) return `${bytes} B`;
+  if (bytes < 1024 * 1024) return `${(bytes / 1024).toFixed(1)} KB`;
+  return `${(bytes / (1024 * 1024)).toFixed(1)} MB`;
+}
+
+function formatDateTime(dt: Date | string | null): string | null {
+  if (!dt) return null;
+  if (dt instanceof Date) {
+    return dt.toLocaleString('ko-KR', {
+      year: 'numeric', month: '2-digit', day: '2-digit',
+      hour: '2-digit', minute: '2-digit', second: '2-digit',
+    });
+  }
+  return String(dt);
+}
+
+interface MetaRowProps {
+  label: string;
+  value: string | null | undefined;
+}
+
+function MetaRow({ label, value }: MetaRowProps) {
+  if (value == null) return null;
+  return (
+    <div className="flex justify-between gap-2 py-0.5 border-b border-border/40 last:border-0 font-korean">
+      <span className="text-text-3 shrink-0">{label}</span>
+      <span className="text-text-1 text-right break-all">{value}</span>
+    </div>
+  );
 }
 
 export function OilAreaAnalysis() {
   const [selectedFiles, setSelectedFiles] = useState<File[]>([]);
   const [previewUrls, setPreviewUrls] = useState<string[]>([]);
-  const [imageGpsInfos, setImageGpsInfos] = useState<(GpsInfo | undefined)[]>([]);
+  const [imageExifs, setImageExifs] = useState<(ImageExif | undefined)[]>([]);
+  const [selectedImageIndex, setSelectedImageIndex] = useState<number | null>(null);
   const [stitchedBlob, setStitchedBlob] = useState<Blob | null>(null);
   const [stitchedPreviewUrl, setStitchedPreviewUrl] = useState<string | null>(null);
   const [isStitching, setIsStitching] = useState(false);
@@ -34,28 +78,45 @@ export function OilAreaAnalysis() {
     // eslint-disable-next-line react-hooks/exhaustive-deps
   }, []);
 
-  // 선택된 파일이 바뀔 때마다 새 파일의 EXIF GPS 추출
+  // 선택된 파일이 바뀔 때마다 새 파일의 EXIF 전체 추출
   useEffect(() => {
     selectedFiles.forEach((file, i) => {
       if (processedFilesRef.current.has(file)) return;
       processedFilesRef.current.add(file);
 
-      exifr.gps(file)
-        .then(gps => {
-          setImageGpsInfos(prev => {
+      exifr.parse(file, { gps: true, exif: true, ifd0: true, translateValues: false })
+        .then(exif => {
+          const info: ImageExif = {
+            lat: exif?.latitude ?? null,
+            lon: exif?.longitude ?? null,
+            altitude: exif?.GPSAltitude ?? null,
+            make: exif?.Make ?? null,
+            model: exif?.Model ?? null,
+            dateTime: exif?.DateTimeOriginal ?? null,
+            exposureTime: exif?.ExposureTime ?? null,
+            fNumber: exif?.FNumber ?? null,
+            iso: exif?.ISO ?? null,
+            focalLength: exif?.FocalLength ?? null,
+            imageWidth: exif?.ImageWidth ?? exif?.ExifImageWidth ?? null,
+            imageHeight: exif?.ImageHeight ?? exif?.ExifImageHeight ?? null,
+          };
+          setImageExifs(prev => {
             const updated = [...prev];
             while (updated.length <= i) updated.push(undefined);
-            updated[i] = gps
-              ? { lat: gps.latitude, lon: gps.longitude }
-              : { lat: null, lon: null };
+            updated[i] = info;
             return updated;
           });
         })
         .catch(() => {
-          setImageGpsInfos(prev => {
+          setImageExifs(prev => {
             const updated = [...prev];
             while (updated.length <= i) updated.push(undefined);
-            updated[i] = { lat: null, lon: null };
+            updated[i] = {
+              lat: null, lon: null, altitude: null,
+              make: null, model: null, dateTime: null,
+              exposureTime: null, fNumber: null, iso: null,
+              focalLength: null, imageWidth: null, imageHeight: null,
+            };
             return updated;
           });
         });
@@ -93,7 +154,13 @@ export function OilAreaAnalysis() {
       URL.revokeObjectURL(prev[idx]);
       return prev.filter((_, i) => i !== idx);
     });
-    setImageGpsInfos(prev => prev.filter((_, i) => i !== idx));
+    setImageExifs(prev => prev.filter((_, i) => i !== idx));
+    setSelectedImageIndex(prev => {
+      if (prev === null) return null;
+      if (prev === idx) return null;
+      if (prev > idx) return prev - 1;
+      return prev;
+    });
     // 합성 결과 초기화 (선택 파일이 바뀌었으므로)
     setStitchedBlob(null);
     if (stitchedPreviewUrl) {
@@ -180,21 +247,74 @@ export function OilAreaAnalysis() {
             <div className="text-[11px] font-bold mb-1.5 font-korean">선택된 이미지</div>
             <div className="flex flex-col gap-1 mb-3">
               {selectedFiles.map((file, i) => (
-                <div
-                  key={`${file.name}-${i}`}
-                  className="flex items-center gap-2 px-2 py-1.5 bg-bg-3 border border-border rounded-sm text-[11px] font-korean"
-                >
-                  <span className="text-primary-cyan">📷</span>
-                  <span className="flex-1 truncate text-text-1">{file.name}</span>
-                  <button
-                    onClick={() => handleRemoveFile(i)}
-                    disabled={isStitching || isAnalyzing}
-                    className="text-text-3 hover:text-status-red transition-colors cursor-pointer
-                               disabled:opacity-40 disabled:cursor-not-allowed ml-1 shrink-0"
-                    title="제거"
+                <div key={`${file.name}-${i}`}>
+                  <div
+                    className={`flex items-center gap-2 px-2 py-1.5 bg-bg-3 border rounded-sm text-[11px] font-korean cursor-pointer transition-colors
+                      ${selectedImageIndex === i ? 'border-primary-cyan' : 'border-border'}`}
+                    onClick={() => setSelectedImageIndex(i)}
                   >
-                    ✕
-                  </button>
+                    <span className="text-primary-cyan">📷</span>
+                    <span className="flex-1 truncate text-text-1">{file.name}</span>
+                    <button
+                      onClick={e => { e.stopPropagation(); handleRemoveFile(i); }}
+                      disabled={isStitching || isAnalyzing}
+                      className="text-text-3 hover:text-status-red transition-colors cursor-pointer
+                                 disabled:opacity-40 disabled:cursor-not-allowed ml-1 shrink-0"
+                      title="제거"
+                    >
+                      ✕
+                    </button>
+                  </div>
+                  {selectedImageIndex === i && imageExifs[i] !== undefined && (
+                    <div className="mt-1 mb-1 px-2 py-1.5 bg-bg-0 border border-border/60 rounded-sm text-[11px] font-korean">
+                      <MetaRow label="파일 크기" value={formatFileSize(file.size)} />
+                      <MetaRow
+                        label="해상도"
+                        value={imageExifs[i]!.imageWidth && imageExifs[i]!.imageHeight
+                          ? `${imageExifs[i]!.imageWidth} × ${imageExifs[i]!.imageHeight}`
+                          : null}
+                      />
+                      <MetaRow label="촬영일시" value={formatDateTime(imageExifs[i]!.dateTime)} />
+                      <MetaRow
+                        label="장비"
+                        value={imageExifs[i]!.make || imageExifs[i]!.model
+                          ? [imageExifs[i]!.make, imageExifs[i]!.model].filter(Boolean).join(' ')
+                          : null}
+                      />
+                      <MetaRow
+                        label="위도"
+                        value={imageExifs[i]!.lat !== null ? decimalToDMS(imageExifs[i]!.lat!, true) : null}
+                      />
+                      <MetaRow
+                        label="경도"
+                        value={imageExifs[i]!.lon !== null ? decimalToDMS(imageExifs[i]!.lon!, false) : null}
+                      />
+                      <MetaRow
+                        label="고도"
+                        value={imageExifs[i]!.altitude !== null ? `${imageExifs[i]!.altitude!.toFixed(1)} m` : null}
+                      />
+                      <MetaRow
+                        label="셔터속도"
+                        value={imageExifs[i]!.exposureTime
+                          ? imageExifs[i]!.exposureTime! < 1
+                            ? `1/${Math.round(1 / imageExifs[i]!.exposureTime!)}s`
+                            : `${imageExifs[i]!.exposureTime}s`
+                          : null}
+                      />
+                      <MetaRow
+                        label="조리개"
+                        value={imageExifs[i]!.fNumber ? `f/${imageExifs[i]!.fNumber}` : null}
+                      />
+                      <MetaRow
+                        label="ISO"
+                        value={imageExifs[i]!.iso ? String(imageExifs[i]!.iso) : null}
+                      />
+                      <MetaRow
+                        label="초점거리"
+                        value={imageExifs[i]!.focalLength ? `${imageExifs[i]!.focalLength} mm` : null}
+                      />
+                    </div>
+                  )}
                 </div>
               ))}
             </div>
@@ -239,8 +359,11 @@ export function OilAreaAnalysis() {
           {Array.from({ length: MAX_IMAGES }).map((_, i) => (
             <div
               key={i}
-              className="bg-bg-3 border border-border rounded-sm overflow-hidden flex flex-col"
+              className={`bg-bg-3 border rounded-sm overflow-hidden flex flex-col transition-colors
+                ${previewUrls[i] ? 'cursor-pointer' : ''}
+                ${selectedImageIndex === i ? 'border-primary-cyan' : 'border-border'}`}
               style={{ height: '300px' }}
+              onClick={() => { if (previewUrls[i]) setSelectedImageIndex(i); }}
             >
               {previewUrls[i] ? (
                 <>
@@ -255,12 +378,12 @@ export function OilAreaAnalysis() {
                     <div className="text-[10px] text-text-2 truncate font-korean flex-1 min-w-0">
                       {selectedFiles[i]?.name}
                     </div>
-                    {imageGpsInfos[i] === undefined ? (
+                    {imageExifs[i] === undefined ? (
                       <div className="text-[10px] text-text-3 font-korean shrink-0">GPS 읽는 중...</div>
-                    ) : imageGpsInfos[i]?.lat !== null ? (
+                    ) : imageExifs[i]?.lat !== null ? (
                       <div className="text-[10px] text-primary-cyan font-mono leading-tight text-right shrink-0">
-                        {decimalToDMS(imageGpsInfos[i]!.lat!, true)}<br />
-                        {decimalToDMS(imageGpsInfos[i]!.lon!, false)}
+                        {decimalToDMS(imageExifs[i]!.lat!, true)}<br />
+                        {decimalToDMS(imageExifs[i]!.lon!, false)}
                       </div>
                     ) : (
                       <div className="text-[10px] text-text-3 font-korean shrink-0">GPS 정보 없음</div>

frontend/src/tabs/prediction/components/LeftPanel.tsx

@@ -20,6 +20,9 @@ export function LeftPanel({
   isRunningSimulation,
   selectedModels,
   onModelsChange,
+  visibleModels,
+  onVisibleModelsChange,
+  hasResults,
   predictionTime,
   onPredictionTimeChange,
   spillType,
@@ -82,6 +85,9 @@ export function LeftPanel({
           isRunningSimulation={isRunningSimulation}
           selectedModels={selectedModels}
           onModelsChange={onModelsChange}
+          visibleModels={visibleModels}
+          onVisibleModelsChange={onVisibleModelsChange}
+          hasResults={hasResults}
           predictionTime={predictionTime}
           onPredictionTimeChange={onPredictionTimeChange}
           spillType={spillType}

frontend/src/tabs/prediction/components/OilSpillView.tsx

@@ -14,7 +14,8 @@ import type { BacktrackPhase, BacktrackVessel, BacktrackConditions, ReplayShip,
 import { TOTAL_REPLAY_FRAMES } from '@common/types/backtrack'
 import { fetchBacktrackByAcdnt, createBacktrack, fetchPredictionDetail, fetchAnalysisTrajectory } from '../services/predictionApi'
 import type { CenterPoint, HydrDataStep, ImageAnalyzeResult, OilParticle, PredictionDetail, SimulationRunResponse, SimulationSummary, WindPoint } from '../services/predictionApi'
-import { useSimulationStatus } from '../hooks/useSimulationStatus'
+import { useMultiSimulationStatus } from '../hooks/useSimulationStatus'
+import type { ModelExecRef } from '../hooks/useSimulationStatus'
 import SimulationLoadingOverlay from './SimulationLoadingOverlay'
 import SimulationErrorModal from './SimulationErrorModal'
 import { api } from '@common/services/api'
@@ -119,11 +120,13 @@ export function OilSpillView() {
   const [isSelectingLocation, setIsSelectingLocation] = useState(false)
   const [oilTrajectory, setOilTrajectory] = useState<OilParticle[]>([])
   const [centerPoints, setCenterPoints] = useState<CenterPoint[]>([])
-  const [windData, setWindData] = useState<WindPoint[][]>([])
-  const [hydrData, setHydrData] = useState<(HydrDataStep | null)[]>([])
+  const [windDataByModel, setWindDataByModel] = useState<Record<string, WindPoint[][]>>({})
+  const [hydrDataByModel, setHydrDataByModel] = useState<Record<string, (HydrDataStep | null)[]>>({})
+  const [windHydrModel, setWindHydrModel] = useState<string>('OpenDrift')
   const [isRunningSimulation, setIsRunningSimulation] = useState(false)
   const [simulationError, setSimulationError] = useState<string | null>(null)
   const [selectedModels, setSelectedModels] = useState<Set<PredictionModel>>(new Set(['OpenDrift']))
+  const [visibleModels, setVisibleModels] = useState<Set<PredictionModel>>(new Set(['OpenDrift']))
   const [predictionTime, setPredictionTime] = useState(48)
   const [accidentTime, setAccidentTime] = useState<string>('')
   const [spillType, setSpillType] = useState('연속')
@@ -154,7 +157,7 @@ export function OilSpillView() {
   // 표시 정보 제어
   const [displayControls, setDisplayControls] = useState<DisplayControls>({
     showCurrent: true,
-    showWind: true,
+    showWind: false,
     showBeached: false,
     showTimeLabel: false,
   })
@@ -188,9 +191,9 @@ export function OilSpillView() {
 
   // 재계산 상태
   const [recalcModalOpen, setRecalcModalOpen] = useState(false)
-  const [currentExecSn, setCurrentExecSn] = useState<number | null>(null)
+  const [pendingExecSns, setPendingExecSns] = useState<ModelExecRef[]>([])
   const [simulationSummary, setSimulationSummary] = useState<SimulationSummary | null>(null)
-  const { data: simStatus } = useSimulationStatus(currentExecSn)
+  const { allDone: simAllDone, anyError: simAnyError, results: simResults, errors: simErrors } = useMultiSimulationStatus(pendingExecSns)
 
   // 오염분석 상태
   const [analysisTab, setAnalysisTab] = useState<'polygon' | 'circle'>('polygon')
@@ -389,34 +392,82 @@ export function OilSpillView() {
     }
   }, [])
 
-  // 시뮬레이션 폴링 결과 처리
+  // 시뮬레이션 폴링 결과 처리 (다중 모델)
   useEffect(() => {
-    if (!simStatus) return;
-    if (simStatus.status === 'DONE' && simStatus.trajectory) {
-      // eslint-disable-next-line react-hooks/set-state-in-effect
-      setOilTrajectory(simStatus.trajectory);
-      setSimulationSummary(simStatus.summary ?? null);
-      setCenterPoints(simStatus.centerPoints ?? []);
-      setWindData(simStatus.windData ?? []);
-      setHydrData(simStatus.hydrData ?? []);
-      setIsRunningSimulation(false);
-      setCurrentExecSn(null);
-      // AI 방어선 자동 생성
-      if (incidentCoord) {
-        const booms = generateAIBoomLines(simStatus.trajectory, incidentCoord, algorithmSettings);
-        setBoomLines(booms);
+    if (pendingExecSns.length === 0) return;
+
+    if (simAllDone) {
+      // 모든 모델의 trajectory 병합 (model 필드 포함)
+      const merged: OilParticle[] = [];
+      let latestSummary: SimulationSummary | null = null;
+      let latestCenterPoints: CenterPoint[] = [];
+      const newWindDataByModel: Record<string, WindPoint[][]> = {};
+      const newHydrDataByModel: Record<string, (HydrDataStep | null)[]> = {};
+
+      simResults.forEach((statusData, model) => {
+        if (statusData.trajectory) {
+          const withModel = statusData.trajectory.map(p => ({ ...p, model }));
+          merged.push(...withModel);
+        }
+        // summary는 OpenDrift 우선, 없으면 다른 모델
+        if (model === 'OpenDrift' || !latestSummary) {
+          if (statusData.summary) latestSummary = statusData.summary;
+        }
+        // windData/hydrData는 모델별로 저장
+        if (statusData.windData) newWindDataByModel[model] = statusData.windData;
+        if (statusData.hydrData) newHydrDataByModel[model] = statusData.hydrData;
+        // centerPoints는 모든 모델 누적 (model 필드 포함 보장)
+        if (statusData.centerPoints) {
+          const withModel = statusData.centerPoints.map(p => ({ ...p, model }));
+          latestCenterPoints = [...latestCenterPoints, ...withModel];
+        }
+      });
+
+      if (merged.length > 0) {
+        setOilTrajectory(merged);
+        const doneModels = new Set<PredictionModel>(
+          Array.from(simResults.entries())
+            .filter(([, s]) => s.trajectory && s.trajectory.length > 0)
+            .map(([m]) => m as PredictionModel)
+        )
+        setVisibleModels(doneModels)
+        setSimulationSummary(latestSummary);
+        setCenterPoints(latestCenterPoints);
+
+        // 데이터가 없는 모델에 OpenDrift(또는 첫 번째 보유 모델) 데이터 복사
+        const refWindData = newWindDataByModel['OpenDrift'] ?? Object.values(newWindDataByModel)[0];
+        const refHydrData = newHydrDataByModel['OpenDrift'] ?? Object.values(newHydrDataByModel)[0];
+        doneModels.forEach(model => {
+          if (!newWindDataByModel[model] && refWindData) newWindDataByModel[model] = refWindData;
+          if (!newHydrDataByModel[model] && refHydrData) newHydrDataByModel[model] = refHydrData;
+        });
+
+        setWindDataByModel(newWindDataByModel);
+        setHydrDataByModel(newHydrDataByModel);
+        setWindHydrModel('OpenDrift');
+        if (incidentCoord) {
+          const booms = generateAIBoomLines(merged, incidentCoord, algorithmSettings);
+          setBoomLines(booms);
+        }
+        setSensitiveResources(DEMO_SENSITIVE_RESOURCES);
+        setCurrentStep(0);
+        setIsPlaying(true);
+        if (incidentCoord) {
+          setFlyToCoord({ lon: incidentCoord.lon, lat: incidentCoord.lat });
+        }
       }
-      setSensitiveResources(DEMO_SENSITIVE_RESOURCES);
-      // 새 시뮬레이션 완료 시 flyTo 없으므로 즉시 재생
-      setCurrentStep(0);
-      setIsPlaying(true);
-    }
-    if (simStatus.status === 'ERROR') {
       setIsRunningSimulation(false);
-      setCurrentExecSn(null);
-      setSimulationError(simStatus.error ?? '시뮬레이션 처리 중 오류가 발생했습니다.');
+      setPendingExecSns([]);
     }
-  }, [simStatus, incidentCoord, algorithmSettings]);
+
+    if (simAnyError) {
+      setIsRunningSimulation(false);
+      setPendingExecSns([]);
+      const errorMessages = Array.from(simErrors.values()).join('; ');
+      setSimulationError(errorMessages || '시뮬레이션 처리 중 오류가 발생했습니다.');
+    }
+  // eslint-disable-next-line react-hooks/exhaustive-deps
+  }, [simAllDone, simAnyError, simResults, simErrors, pendingExecSns.length, incidentCoord, algorithmSettings]);
 
   // trajectory 변경 시 플레이어 스텝 초기화 (재생은 각 경로에서 별도 처리)
   useEffect(() => {
@@ -435,6 +486,17 @@ export function OilSpillView() {
 
   const maxTime = timeSteps[timeSteps.length - 1] ?? predictionTime;
 
+  // 유향유속/풍향풍속 데이터 — 선택한 모델 기준으로 파생
+  const windHydrModelOptions = useMemo(() => Array.from(visibleModels), [visibleModels])
+  const windData = useMemo(
+    () => windDataByModel[windHydrModel] ?? [],
+    [windDataByModel, windHydrModel]
+  )
+  const hydrData = useMemo(
+    () => hydrDataByModel[windHydrModel] ?? [],
+    [hydrDataByModel, windHydrModel]
+  )
+
   useEffect(() => {
     if (!isPlaying || timeSteps.length === 0) return;
     if (currentStep >= maxTime) {
@@ -482,6 +544,7 @@ export function OilSpillView() {
     if (analysis.poseidonStatus !== 'pending') models.add('POSEIDON')
     if (analysis.opendriftStatus !== 'pending') models.add('OpenDrift')
     setSelectedModels(models)
+    setVisibleModels(models)
     // 분석 상세 로딩 (선박/기상 정보)
     try {
       const detail = await fetchPredictionDetail(analysis.acdntSn)
@@ -505,8 +568,9 @@ export function OilSpillView() {
           setOilTrajectory(trajectory)
           if (summary) setSimulationSummary(summary)
           setCenterPoints(cp ?? [])
-          setWindData(wd ?? [])
-          setHydrData(hd ?? [])
+          setWindDataByModel(wd && wd.length > 0 ? { 'OpenDrift': wd } : {})
+          setHydrDataByModel(hd && hd.length > 0 ? { 'OpenDrift': hd } : {})
+          setWindHydrModel('OpenDrift')
           if (coord) setBoomLines(generateAIBoomLines(trajectory, coord, algorithmSettings))
           setSensitiveResources(DEMO_SENSITIVE_RESOURCES)
           // incidentCoord가 변경된 경우 flyTo 완료 후 재생, 그렇지 않으면 즉시 재생
@@ -622,6 +686,7 @@ export function OilSpillView() {
       backtrackStatus: 'pending',
       analyst: '',
       officeName: '',
+      acdntSttsCd: 'ACTIVE',
     })
   }, [])
 
@@ -663,8 +728,12 @@ export function OilSpillView() {
         payload.spillTypeCd = spillType;
       }
 
+      payload.models = Array.from(selectedModels);
+
       const { data } = await api.post<SimulationRunResponse>('/simulation/run', payload);
-      setCurrentExecSn(data.execSn);
+      setPendingExecSns(
+        data.execSns ?? (data.execSn ? [{ model: 'OpenDrift', execSn: data.execSn }] : [])
+      );
 
       // 직접 입력으로 신규 생성된 경우: selectedAnalysis 갱신 + incidentName 초기화
       if (data.acdntSn && isDirectInput) {
@@ -782,6 +851,9 @@ export function OilSpillView() {
           isRunningSimulation={isRunningSimulation}
           selectedModels={selectedModels}
           onModelsChange={setSelectedModels}
+          visibleModels={visibleModels}
+          onVisibleModelsChange={setVisibleModels}
+          hasResults={oilTrajectory.length > 0}
           predictionTime={predictionTime}
           onPredictionTimeChange={setPredictionTime}
           spillType={spillType}
@@ -829,7 +901,7 @@ export function OilSpillView() {
             flyToIncident={flyToCoord}
             isSelectingLocation={isSelectingLocation || isDrawingBoom || drawAnalysisMode === 'polygon'}
             onMapClick={handleMapClick}
-            oilTrajectory={oilTrajectory}
+            oilTrajectory={oilTrajectory.filter(p => visibleModels.has((p.model || 'OpenDrift') as PredictionModel))}
             selectedModels={selectedModels}
             boomLines={boomLines}
             isDrawingBoom={isDrawingBoom}
@@ -838,7 +910,7 @@ export function OilSpillView() {
             layerBrightness={layerBrightness}
             sensitiveResources={sensitiveResources}
             lightMode
-            centerPoints={centerPoints}
+            centerPoints={centerPoints.filter(p => visibleModels.has((p.model || 'OpenDrift') as PredictionModel))}
             windData={windData}
             hydrData={hydrData}
             flyToTarget={flyToTarget}
@@ -1055,6 +1127,9 @@ export function OilSpillView() {
           summary={simulationSummary}
           displayControls={displayControls}
           onDisplayControlsChange={setDisplayControls}
+          windHydrModel={windHydrModel}
+          windHydrModelOptions={windHydrModelOptions}
+          onWindHydrModelChange={setWindHydrModel}
           analysisTab={analysisTab}
           onSwitchAnalysisTab={setAnalysisTab}
           drawAnalysisMode={drawAnalysisMode}
@@ -1074,8 +1149,8 @@ export function OilSpillView() {
       {/* 확산 예측 실행 중 로딩 오버레이 */}
       {isRunningSimulation && (
         <SimulationLoadingOverlay
-          status={simStatus?.status === 'RUNNING' ? 'RUNNING' : 'PENDING'}
-          progress={simStatus?.progress}
+          status="RUNNING"
+          progress={undefined}
         />
       )}
 

frontend/src/tabs/prediction/components/PredictionInputSection.tsx

@@ -18,6 +18,9 @@ interface PredictionInputSectionProps {
   isRunningSimulation: boolean
   selectedModels: Set<PredictionModel>
   onModelsChange: (models: Set<PredictionModel>) => void
+  visibleModels?: Set<PredictionModel>
+  onVisibleModelsChange?: (models: Set<PredictionModel>) => void
+  hasResults?: boolean
   predictionTime: number
   onPredictionTimeChange: (time: number) => void
   spillType: string
@@ -46,6 +49,9 @@ const PredictionInputSection = ({
   isRunningSimulation,
   selectedModels,
   onModelsChange,
+  visibleModels,
+  onVisibleModelsChange,
+  hasResults,
   predictionTime,
   onPredictionTimeChange,
   spillType,
@@ -387,19 +393,21 @@ const PredictionInputSection = ({
             ] as const).map(m => (
               <div
                 key={m.id}
-                className={`prd-mc ${selectedModels.has(m.id) ? 'on' : ''} cursor-pointer`}
+                className={`prd-mc ${(hasResults ? (visibleModels ?? selectedModels) : selectedModels).has(m.id) ? 'on' : ''} cursor-pointer`}
                 onClick={() => {
                   if (!m.ready) {
                     alert(`${m.id} 모델은 현재 준비중입니다.`)
                     return
                   }
-                  const next = new Set(selectedModels)
-                  if (next.has(m.id)) {
-                    next.delete(m.id)
+                  if (hasResults && onVisibleModelsChange) {
+                    const next = new Set(visibleModels ?? selectedModels)
+                    if (next.has(m.id)) { next.delete(m.id) } else { next.add(m.id) }
+                    onVisibleModelsChange(next)
                   } else {
-                    next.add(m.id)
+                    const next = new Set(selectedModels)
+                    if (next.has(m.id)) { next.delete(m.id) } else { next.add(m.id) }
+                    onModelsChange(next)
                   }
-                  onModelsChange(next)
                 }}
               >
                 <span className="prd-md" style={{ background: m.color }} />

frontend/src/tabs/prediction/components/RightPanel.tsx

@@ -18,6 +18,9 @@ interface RightPanelProps {
   summary?: SimulationSummary | null
   displayControls?: DisplayControls
   onDisplayControlsChange?: (controls: DisplayControls) => void
+  windHydrModel?: string
+  windHydrModelOptions?: string[]
+  onWindHydrModelChange?: (model: string) => void
   analysisTab?: 'polygon' | 'circle'
   onSwitchAnalysisTab?: (tab: 'polygon' | 'circle') => void
   drawAnalysisMode?: 'polygon' | null
@@ -36,6 +39,7 @@ interface RightPanelProps {
 export function RightPanel({
   onOpenBacktrack, onOpenRecalc, onOpenReport, detail, summary,
   displayControls, onDisplayControlsChange,
+  windHydrModel, windHydrModelOptions = [], onWindHydrModelChange,
   analysisTab = 'polygon', onSwitchAnalysisTab,
   drawAnalysisMode, analysisPolygonPoints = [],
   circleRadiusNm = 5, onCircleRadiusChange,
@@ -85,6 +89,20 @@ export function RightPanel({
               onChange={(v) => onDisplayControlsChange?.({ ...displayControls!, showTimeLabel: v })}
             >시간 표시</ControlledCheckbox>
           </div>
+          {windHydrModelOptions.length > 1 && (
+            <div className="flex items-center gap-2 mt-1.5">
+              <span className="text-[9px] text-text-3 font-korean whitespace-nowrap">데이터 모델</span>
+              <select
+                value={windHydrModel}
+                onChange={e => onWindHydrModelChange?.(e.target.value)}
+                className="flex-1 text-[9px] bg-bg-3 border border-border rounded px-1 py-0.5 text-text-2 font-korean"
+              >
+                {windHydrModelOptions.map(m => (
+                  <option key={m} value={m}>{m}</option>
+                ))}
+              </select>
+            </div>
+          )}
         </Section>
 
         {/* 오염분석 */}

frontend/src/tabs/prediction/components/leftPanelTypes.ts

@@ -17,6 +17,9 @@ export interface LeftPanelProps {
   isRunningSimulation: boolean
   selectedModels: Set<PredictionModel>
   onModelsChange: (models: Set<PredictionModel>) => void
+  visibleModels?: Set<PredictionModel>
+  onVisibleModelsChange?: (models: Set<PredictionModel>) => void
+  hasResults?: boolean
   predictionTime: number
   onPredictionTimeChange: (time: number) => void
   spillType: string

frontend/src/tabs/prediction/hooks/useSimulationStatus.ts

@@ -1,4 +1,4 @@
-import { useQuery } from '@tanstack/react-query';
+import { useQuery, useQueries } from '@tanstack/react-query';
 import { api } from '@common/services/api';
 import type { SimulationStatusResponse } from '../services/predictionApi';
 
@@ -14,3 +14,73 @@ export const useSimulationStatus = (execSn: number | null) => {
     },
   });
 };
+
+export interface ModelExecRef {
+  model: string;
+  execSn: number;
+}
+
+interface MultiSimulationStatus {
+  allDone: boolean;
+  anyError: boolean;
+  isLoading: boolean;
+  results: Map<string, SimulationStatusResponse>;
+  errors: Map<string, string>;
+}
+
+export const useMultiSimulationStatus = (execSns: ModelExecRef[]): MultiSimulationStatus => {
+  const queries = useQueries({
+    queries: execSns.map(({ model, execSn }) => ({
+      queryKey: ['simulationStatus', execSn],
+      queryFn: () =>
+        api.get<SimulationStatusResponse>(`/simulation/status/${execSn}`).then(r => r.data),
+      enabled: execSns.length > 0,
+      refetchInterval: (query: { state: { data?: SimulationStatusResponse } }) => {
+        const status = query.state.data?.status;
+        if (status === 'DONE' || status === 'ERROR') return false;
+        return 3000;
+      },
+      meta: { model },
+    })),
+  });
+
+  if (execSns.length === 0) {
+    return {
+      allDone: false,
+      anyError: false,
+      isLoading: false,
+      results: new Map(),
+      errors: new Map(),
+    };
+  }
+
+  const results = new Map<string, SimulationStatusResponse>();
+  const errors = new Map<string, string>();
+
+  execSns.forEach(({ model }, index) => {
+    const query = queries[index];
+    if (query.data) {
+      results.set(model, query.data);
+    }
+    if (query.error) {
+      const err = query.error;
+      errors.set(model, err instanceof Error ? err.message : String(err));
+    }
+  });
+
+  const allDone =
+    execSns.length > 0 &&
+    execSns.every((_, index) => {
+      const status = queries[index].data?.status;
+      return status === 'DONE' || status === 'ERROR';
+    });
+
+  const anyError = execSns.some((_, index) => {
+    const status = queries[index].data?.status;
+    return status === 'ERROR' || queries[index].isError;
+  });
+
+  const isLoading = execSns.some((_, index) => queries[index].isLoading);
+
+  return { allDone, anyError, isLoading, results, errors };
+};

frontend/src/tabs/prediction/services/predictionApi.ts

@@ -123,7 +123,8 @@ export const createBacktrack = async (input: {
 
 export interface SimulationRunResponse {
   success: boolean;
-  execSn: number;
+  execSn: number;  // 하위 호환 유지 (첫 번째 모델의 execSn)
+  execSns: Array<{ model: string; execSn: number }>;
   acdntSn: number | null;
   status: 'RUNNING';
 }
@@ -152,6 +153,7 @@ export interface CenterPoint {
   lat: number;
   lon: number;
   time: number;
+  model?: string;
 }
 
 export interface OilParticle {
@@ -160,6 +162,7 @@ export interface OilParticle {
   time: number;
   particle?: number;
   stranded?: 0 | 1;
+  model?: string;
 }
 
 export interface SimulationSummary {