sync embedded/arc_tz_weather from arc_tz_weather

Author: actionresearchprojects

embedded/arc_tz_weather/index.html

@@ -28,6 +28,9 @@
 KPH_TO_KN = 250 / 463   # knots per km/h
 MS_TO_KN  = 900 / 463   # knots per m/s
 
+# Calm threshold: 0.1 m/s expressed in km/h (0.1 * 3.6 = 0.36)
+CALM_THRESHOLD_KPH = 0.36
+
 # Thresholds in knots (WMO integer boundaries). range_kn is [lo, hi).
 BEAUFORT_SCALE = {
     0:  {"range_kn": (0,   1),   "label": "Calm"},
@@ -368,6 +371,14 @@ def get_season(month):
     return "Unknown"
 
 
+_SEASON_BOUNDARY_LABELS = {
+    1:  "January Dry Season (Kiangazi)",
+    3:  "Long Rains (Masika)",
+    6:  "June Dry Season (Kiangazi)",
+    11: "Short Rains (Vuli)",
+}
+
+
 def get_season_boundaries(df):
     """Get season boundary timestamps for marking on charts."""
     boundaries = []
@@ -383,8 +394,7 @@ def get_season_boundaries(df):
         for m in season_starts:
             dt = TIMEZONE.localize(datetime(year, m, 1))
             if min_date <= dt <= max_date:
-                season_name = get_season(m)
-                boundaries.append({"ts": to_eat_ms(dt), "label": season_name})
+                boundaries.append({"ts": to_eat_ms(dt), "label": _SEASON_BOUNDARY_LABELS[m]})
         year += 1
     return boundaries
 

embedded/arc_tz_weather/modules/common.py

@@ -10,7 +10,7 @@
 import numpy as np
 
 from .common import (
-    VENTILATION_COLORS, to_eat_ms, compass_bin,
+    VENTILATION_COLORS, CALM_THRESHOLD_KPH, to_eat_ms, compass_bin,
     get_season_boundaries,
 )
 
@@ -59,7 +59,7 @@ def process(df, rain_events=None):
 def _build_driving_rain_index(xdf):
     """Build Driving Rain Index polar chart and time series."""
     # Only readings with both wind and rain
-    wr = xdf[(xdf["avg_wind_kph"] > 0) & (xdf["precip_rate_mmh"] > 0)].copy()
+    wr = xdf[(xdf["avg_wind_kph"] > CALM_THRESHOLD_KPH) & (xdf["precip_rate_mmh"] > 0)].copy()
 
     if len(wr) == 0:
         return {"id": "driving-rain", "title": "Driving Rain Index",

embedded/arc_tz_weather/modules/cross_variable.py

@@ -301,14 +301,14 @@ def _build_intensity_distribution(pdf):
 
 
 def _build_diurnal_rainfall(pdf):
-    """Build diurnal rainfall pattern: mean hourly rainfall + rain probability."""
+    """Build diurnal rainfall pattern: mean hourly rainfall + rain frequency."""
     pdf_c = pdf.copy()
     pdf_c["hour"] = pdf_c["timestamp"].dt.hour
 
     # Mean rainfall per hour (from incremental)
     hourly_rain = pdf_c.groupby("hour")["precip_incr"].mean()
 
-    # Rain probability by hour
+    # Rain frequency by hour
     hourly_prob = pdf_c.groupby("hour").apply(
         lambda g: (g["precip_rate_mmh"] > 0).sum() / len(g) * 100
     )
@@ -328,7 +328,7 @@ def _build_diurnal_rainfall(pdf):
         {
             "type": "scatter",
             "mode": "lines+markers",
-            "name": "Rain Probability (%)",
+            "name": "Rain Frequency (%)",
             "x": hours,
             "y": rain_probs,
             "yaxis": "y2",
@@ -341,7 +341,7 @@ def _build_diurnal_rainfall(pdf):
         "xaxis": {"title": "Hour of Day (EAT)", "dtick": 1},
         "yaxis": {"title": "Mean Rainfall (mm)"},
         "yaxis2": {
-            "title": "Rain Probability (%)",
+            "title": "Rain Frequency (%)",
             "overlaying": "y",
             "side": "right",
             "range": [0, max(rain_probs) * 1.2] if rain_probs else [0, 100],

embedded/arc_tz_weather/modules/precipitation.py

@@ -12,7 +12,7 @@
 from .common import (
     COMPASS_DIRS_16, COMPASS_DIRS_8, WIND_SPEED_BINS, WIND_SPEED_LABELS,
     WIND_SPEED_COLORS, BEAUFORT_SCALE, VENTILATION_COLORS, WIND_CLASSIFICATIONS,
-    KN_TO_KPH, TIMEZONE,
+    KN_TO_KPH, TIMEZONE, CALM_THRESHOLD_KPH,
     spike_filter, compass_bin, beaufort_number, weibull_fit, to_eat_ms,
     get_season_boundaries,
 )
@@ -40,10 +40,10 @@ def process(df):
 
     # ── Summary Statistics ────────────────────────────────────────────────
     total_readings = len(wdf)
-    calm_readings = (wdf["avg_wind_kph"] == 0).sum()
+    calm_readings = (wdf["avg_wind_kph"] <= CALM_THRESHOLD_KPH).sum()
     calm_pct = round(calm_readings / total_readings * 100, 1) if total_readings else 0
 
-    non_calm = wdf[wdf["avg_wind_kph"] > 0]
+    non_calm = wdf[wdf["avg_wind_kph"] > CALM_THRESHOLD_KPH]
     mean_speed = round(wdf["avg_wind_kph"].mean(), 1)
     mean_speed_noncalm = round(non_calm["avg_wind_kph"].mean(), 1) if len(non_calm) else 0
     max_speed = round(wdf["avg_wind_kph"].max(), 1)
@@ -108,9 +108,9 @@ def _build_wind_rose(wdf, n_points):
     dir_labels = [d[0] for d in dirs]
     col = "compass_16" if n_points == 16 else "compass_8"
 
-    non_calm = wdf[wdf["avg_wind_kph"] > 0].copy()
+    non_calm = wdf[wdf["avg_wind_kph"] > CALM_THRESHOLD_KPH].copy()
     total = len(wdf)
-    calm_pct = round((wdf["avg_wind_kph"] == 0).sum() / total * 100, 1) if total else 0
+    calm_pct = round((wdf["avg_wind_kph"] <= CALM_THRESHOLD_KPH).sum() / total * 100, 1) if total else 0
 
     # Speed bins: 0-5, 5-10, 10-15, 15-20, 20+
     traces = []
@@ -176,9 +176,9 @@ def _build_wind_timeseries(wdf):
     """Build wind speed time series with average and gust."""
     timestamps = [to_eat_ms(t) for t in wdf["timestamp"]]
 
-    # 24-hour running mean
+    # 12-hour running mean
     wdf_sorted = wdf.set_index("timestamp").sort_index()
-    running_mean = wdf_sorted["avg_wind_kph"].rolling("24h", min_periods=1).mean()
+    running_mean = wdf_sorted["avg_wind_kph"].rolling("12h", center=True, min_periods=1).mean()
     rm_ts = [to_eat_ms(t) for t in running_mean.index]
     rm_vals = [round(v, 1) if not pd.isna(v) else None for v in running_mean.values]
 
@@ -208,7 +208,7 @@ def _build_wind_timeseries(wdf):
         {
             "type": "scatter",
             "mode": "lines",
-            "name": "24h Mean",
+            "name": "12h Mean",
             "x_ms": rm_ts,
             "y": rm_vals,
             "line": {"color": "#d62728", "width": 2},
@@ -242,7 +242,7 @@ def _build_diurnal_wind(wdf):
 
     # Calm percentage by hour
     calm_by_hour = wdf_c.groupby("hour").apply(
-        lambda g: (g["avg_wind_kph"] == 0).sum() / len(g) * 100
+        lambda g: (g["avg_wind_kph"] <= CALM_THRESHOLD_KPH).sum() / len(g) * 100
     ).round(1)
 
     # Optional: separate by month
@@ -293,7 +293,7 @@ def _build_diurnal_wind(wdf):
         },
         {
             "type": "bar",
-            "name": "Calm %",
+            "name": "Calm % (\u22640.1 m/s)",
             "x": hours,
             "y": calm_pcts,
             "yaxis": "y2",
@@ -334,7 +334,7 @@ def _build_wind_distribution(wdf, k_val, c_val):
     hist, bin_edges = np.histogram(speeds, bins=bins)
 
     # Separate calm bar
-    calm_count = int((speeds == 0).sum())
+    calm_count = int((speeds <= CALM_THRESHOLD_KPH).sum())
     bin_centers = [(bin_edges[i] + bin_edges[i + 1]) / 2 for i in range(len(hist))]
 
     traces = [{
@@ -387,7 +387,7 @@ def _build_wind_distribution(wdf, k_val, c_val):
 
 def _build_gust_factor(wdf):
     """Build gust factor scatter plot (gust factor vs avg speed, colored by hour)."""
-    valid = wdf[(wdf["avg_wind_kph"] > 0) & wdf["peak_wind_kph"].notna()].copy()
+    valid = wdf[(wdf["avg_wind_kph"] > CALM_THRESHOLD_KPH) & wdf["peak_wind_kph"].notna()].copy()
     valid["gust_factor"] = valid["peak_wind_kph"] / valid["avg_wind_kph"]
     valid["hour"] = valid["timestamp"].dt.hour
 
@@ -440,7 +440,7 @@ def _build_gust_factor(wdf):
 
 def _build_calm_periods(wdf):
     """Build calm period analysis: distribution of consecutive calm durations."""
-    is_calm = (wdf["avg_wind_kph"] == 0).values
+    is_calm = (wdf["avg_wind_kph"] <= CALM_THRESHOLD_KPH).values
     timestamps = wdf["timestamp"].values
 
     # Find consecutive calm periods
@@ -528,8 +528,8 @@ def _build_ventilation_availability(wdf):
         for date, group in wdf_c.groupby("date"):
             total = len(group)
             effective = (group["avg_wind_kph"] >= thresh).sum()
-            marginal = ((group["avg_wind_kph"] > 0) & (group["avg_wind_kph"] < thresh)).sum()
-            calm = (group["avg_wind_kph"] == 0).sum()
+            marginal = ((group["avg_wind_kph"] > CALM_THRESHOLD_KPH) & (group["avg_wind_kph"] < thresh)).sum()
+            calm = (group["avg_wind_kph"] <= CALM_THRESHOLD_KPH).sum()
 
             # Convert to hours (assuming ~5-min intervals)
             interval_h = 5 / 60
@@ -545,8 +545,8 @@ def _build_ventilation_availability(wdf):
     # Default threshold stats
     default_thresh = 3.5
     all_effective = (wdf_c["avg_wind_kph"] >= default_thresh).sum()
-    all_marginal = ((wdf_c["avg_wind_kph"] > 0) & (wdf_c["avg_wind_kph"] < default_thresh)).sum()
-    all_calm = (wdf_c["avg_wind_kph"] == 0).sum()
+    all_marginal = ((wdf_c["avg_wind_kph"] > CALM_THRESHOLD_KPH) & (wdf_c["avg_wind_kph"] < default_thresh)).sum()
+    all_calm = (wdf_c["avg_wind_kph"] <= CALM_THRESHOLD_KPH).sum()
     total = len(wdf_c)
     eff_pct = round(all_effective / total * 100, 1) if total else 0
 
@@ -583,7 +583,7 @@ def _build_ventilation_availability(wdf):
         {
             "type": "scatter",
             "mode": "lines",
-            "name": "Calm",
+            "name": "Calm (\u22640.1 m/s)",
             "x_ms": dates_ms,
             "y": calm_hours,
             "fill": "tonexty",