feat: implement multi-source weather data collection for OpenWeatherMap, Visual Crossing, and NOAA METAR.

2026-02-08 02:27:21 +08:00
parent 7e636447ce
commit c065992fb0
2 changed files with 75 additions and 22 deletions
@@ -13,7 +13,7 @@ from src.utils.config_loader import load_config
 from src.data_collection.weather_sources import WeatherDataCollector

 def analyze_weather_trend(weather_data, temp_symbol):
-    """根据实测与预测分析气温态势"""
+    """根据实测与预测分析气温态势，增加峰值时刻预测"""
    insights = []
    
    metar = weather_data.get("metar", {})
@@ -23,46 +23,70 @@ def analyze_weather_trend(weather_data, temp_symbol):
        return ""
        
    curr_temp = metar.get("current", {}).get("temp")
-    forecast_high = open_meteo.get("daily", {}).get("temperature_2m_max", [None])[0]
+    daily = open_meteo.get("daily", {})
+    forecast_high = daily.get("temperature_2m_max", [None])[0]
    wind_speed = metar.get("current", {}).get("wind_speed_kt", 0)
    
    # 获取当地时间小时
-    local_time_str = open_meteo.get("current", {}).get("local_time", "")
+    local_time_full = open_meteo.get("current", {}).get("local_time", "")
    try:
-        local_hour = int(local_time_str.split(" ")[1].split(":")[0])
+        local_date_str = local_time_full.split(" ")[0] # YYYY-MM-DD
+        local_hour = int(local_time_full.split(" ")[1].split(":")[0])
    except:
-        local_hour = datetime.now().hour # 降级方案
+        local_date_str = datetime.now().strftime("%Y-%m-%d")
+        local_hour = datetime.now().hour
+
+    # --- 增加：峰值时刻预测逻辑 ---
+    hourly = open_meteo.get("hourly", {})
+    times = hourly.get("time", [])
+    # 优先寻找高精模型的逐小时数据
+    temps = hourly.get("temperature_2m_hrrr_conus") or hourly.get("temperature_2m_ecmwf_ifs") or hourly.get("temperature_2m", [])
+    
+    peak_hours = []
+    if times and temps and forecast_high is not None:
+        for t_str, temp in zip(times, temps):
+            if t_str.startswith(local_date_str):
+                # 记录所有接近最高温的小时 (容差 0.2)
+                if abs(temp - forecast_high) <= 0.2:
+                    hour = t_str.split("T")[1][:5]
+                    peak_hours.append(hour)
        
+        if peak_hours:
+            window = f"{peak_hours[0]} - {peak_hours[-1]}" if len(peak_hours) > 1 else peak_hours[0]
+            insights.append(f"⏱️ <b>预计峰值时刻</b>：今天 <b>{window}</b> 之间。")
+            if local_hour < int(peak_hours[0].split(":")[0]):
+                insights.append(f"🎯 <b>博弈建议</b>：关注该时段实测能否站稳 {forecast_high}{temp_symbol}。")
+
    if curr_temp is not None and forecast_high is not None:
        diff = forecast_high - curr_temp
        
        # 1. 气温节奏判定
-        if local_hour >= 16:
+        if local_hour >= 17:
            if curr_temp >= forecast_high - 0.5:
-                insights.append(f"✅ <b>今日峰值已达</b>：当前 {curr_temp}{temp_symbol} 已触及预报最高温，后续将进入回落通道。")
+                insights.append(f"✅ <b>今日峰值已达</b>：当前已触及预报最高，大概率已定格。")
            else:
-                insights.append(f"📉 <b>处于降温期</b>：气温已开始从峰值下滑，今日大概率不会再反弹。")
-        elif 11 <= local_hour < 16:
-            if diff > 1.5:
-                insights.append(f"📈 <b>升温进程中</b>：距离预报最高温还有约 {diff:.1f}° 空间，午后余热尚存。")
+                insights.append(f"📉 <b>处于降温期</b>：气温已跌落峰值，今日反弹乏力。")
+        elif 10 <= local_hour < 17:
+            if diff > 1.2:
+                insights.append(f"📈 <b>升温进程中</b>：距离峰值还有约 {diff:.1f}° 空间，正向高点冲击。")
            else:
-                insights.append(f"⚖️ <b>高位横盘</b>：气温已基本涨满，将在当前水平小幅波动，直至日落。")
+                insights.append(f"⚖️ <b>高位横盘</b>：气温已在高位，将在当前水平小幅波动。")
        else:
-            insights.append(f"🌅 <b>早间爬升</b>：气温正在起步。")
+            insights.append(f"🌅 <b>早间爬升</b>：气温正快速起步，等待午后冲击。")

        # 2. 湿度与露点带来的“粘性”分析
        humidity = metar.get("current", {}).get("humidity")
        dewpoint = metar.get("current", {}).get("dewpoint")
        
-        if humidity and humidity > 80:
-            insights.append(f"💦 <b>闷热高湿</b>：空气湿度极大 ({humidity}%)，这会像保温层一样锁住热量，导致夜间降温非常缓慢。")
+        if humidity and humidity > 80 and local_hour >= 18:
+            insights.append(f"💦 <b>闷热高湿</b>：湿度极高 ({humidity}%)，将显著锁住夜间热量。")
        
        if dewpoint is not None and curr_temp - dewpoint < 2.0 and local_hour >= 18:
-            insights.append(f"🌡️ <b>触及露点底线</b>：气温已非常接近露点，进一步下降的空间将被强力压缩，气温将“跌不动了”。")
+            insights.append(f"🌡️ <b>触及露点支撑</b>：气温已跌至露点支撑位，降温将变慢。")

-        # 3. 风力带来的剧烈波动预警
+        # 3. 风力
        if wind_speed >= 15:
-            insights.append(f"🌬️ <b>大风预警 ({wind_speed}kt)</b>：强风可能带来锋面过境，注意气温可能出现非正常的剧烈跳变。")
+            insights.append(f"🌬️ <b>大风预判</b>：当前风力较大 ({wind_speed}kt)，气温可能出现非线性波动。")
        elif wind_speed >= 10:
            insights.append(f"🍃 <b>清劲风</b>：空气流动快，虽然有助于散热，但可能伴随阵风引起微小波动。")

@@ -163,10 +187,22 @@ def start_bot():
            city_today_str = city_now.strftime("%Y-%m-%d")

            msg_lines.append(f"\n📊 <b>Open-Meteo 7天预测</b>")
+            model_split = daily.get("model_split")
            for i, (d, t) in enumerate(zip(dates[:7], max_temps[:7])):
                day_label = "今天" if d == city_today_str else d[5:]
                indicator = "👉 " if d == city_today_str else "   "
-                msg_lines.append(f"{indicator}{day_label}: 最高 {t}{temp_symbol}")
+                
+                # 如果是今天且存在模型分歧，则特别标注
+                if d == city_today_str and model_split:
+                    ecmwf = model_split.get("ecmwf")
+                    hrrr = model_split.get("hrrr")
+                    if ecmwf and hrrr and abs(ecmwf - hrrr) > 0.5:
+                        msg_lines.append(f"{indicator}{day_label}: 最高 {t}{temp_symbol} ⚠️")
+                        msg_lines.append(f"   (模型分歧: ECMWF {ecmwf} | HRRR {hrrr})")
+                    else:
+                        msg_lines.append(f"{indicator}{day_label}: 最高 {t}{temp_symbol}")
+                else:
+                    msg_lines.append(f"{indicator}{day_label}: 最高 {t}{temp_symbol}")

            if metar:
                icao = metar.get("icao", "")
@@ -343,9 +343,10 @@ class WeatherDataCollector:
                "_t": int(time.time()),  # 禁用缓存，强制刷新
            }

-            # 对于美国市场，使用华氏度
+            # 对于美国市场，使用华氏度并请求更多的模型共识
            if use_fahrenheit:
                params["temperature_unit"] = "fahrenheit"
+                params["models"] = "ecmwf_ifs,hrrr_conus"

            response = self.session.get(
                url,
@@ -360,7 +361,23 @@ class WeatherDataCollector:
            utc_offset = data.get("utc_offset_seconds", 0)
            timezone_name = data.get("timezone", "UTC")

-            # 计算精确的当地时间而不是气象站 bucket 时间
+            # 处理多模型数据 (如果请求了 models 参数，返回结构会变化)
+            daily_data = data.get("daily", {})
+            if "temperature_2m_max_ecmwf_ifs" in daily_data:
+                # 获取首日的各模型峰值比较
+                ecmwf_max = daily_data.get("temperature_2m_max_ecmwf_ifs", [])
+                hrrr_max = daily_data.get("temperature_2m_max_hrrr_conus", [])
+                
+                # 记录多模型分歧
+                daily_data["model_split"] = {
+                    "ecmwf": ecmwf_max[0] if ecmwf_max else None,
+                    "hrrr": hrrr_max[0] if hrrr_max else None
+                }
+                # 设置主显示值为 HRRR (当地高精)
+                if hrrr_max:
+                    daily_data["temperature_2m_max"] = hrrr_max
+
+            # 计算精确的当地时间
            now_utc = datetime.utcnow()
            local_now = now_utc + timedelta(seconds=utc_offset)
            local_time_str = local_now.strftime("%Y-%m-%d %H:%M")
@@ -375,7 +392,7 @@ class WeatherDataCollector:
                    "local_time": local_time_str,
                },
                "hourly": data.get("hourly", {}),
-                "daily": data.get("daily", {}),
+                "daily": daily_data,
                "unit": "fahrenheit" if use_fahrenheit else "celsius",
            }
        except Exception as e: