By Paul Homewood
As promised, the puzzle of the England & Wales Rainfall Series.
As should now be clear, the wettest 3-month period, since Met Office records began in 1910, was in 1929, both for the UK as a whole, and for England. However, as Elim Rosie correctly pointed out, for the long running England & Wales Series, October to December 2000 was wetter still. To avoid confusion, I will hereafter refer to this series as the EWP.
(First, a quick recap. The EWP is the series which dates back to 1766. As well as this series, dating back to 1766, the Met Office also produce regional statistics back to 1910, from which the separate England & Wales figures are derived).
Given that England was wetter in 1929, it seemed logical that Wales must have been much, much wetter in 2000. But when I checked, it turns out to be nothing of the sort. As in England, October to December 1929 was considerably wetter than the same months of 2000 in Wales.
Let’s check the figures.
| |
EWP Series |
England |
Wales |
| Oct-Dec 1929 |
499.2 |
452.1 |
851.9 |
| Oct-Dec 2000 |
512.3 |
442.1 |
803.0 |
| Dec-Feb 2014 |
456.1 |
396.8 |
739.1 |
Rainfall mm
As you can see, the three months in both England and Wales were considerably wetter in 1929, in stark contrast to the England & Wales Series.
The
area of England & Wales is 50337 and 8018 sq miles respectively, so we can come up with a weighted average for England & Wales, based on the Met Office data since 1910:
| |
mm |
| Oct-Dec 1929 |
507.1 |
| Oct-Dec 2000 |
491.6 |
| Dec-Feb 2014 |
443.8 |
Whereas the EWP data shows 2000 as being 13.1mm wetter, the Met Office regional stats give 1929 as being 15.5mm wetter than 2000. The difference between the two sets of 28.6mm is sizeable.
When judged against the latter dataset, the EWP Series UNDERESTIMATES rainfall in 1929, and OVERESTIMATES it in 2000. Furthermore, it also overestimates it for last winter, suggesting this is not a one off problem, and that the EWP may have a “wet bias”.
Of course, there are bound to be some slight differences between the two sets, as they are built up differently. In particular, there is a need for the EWP to match and be consistent with the records and methodology employed two centuries ago, when data was more sparse.
Nevertheless, it cannot be claimed that both the Met Office dataset and the EWP Series accurately reflect national patterns and trends.
According to the
Met Office, their figures are built up from 5 x 5 km gridded datasets, so logically ought to be much more accurate than the EWP.
Does any of this matter? Is it just an academic matter?
Well, arguably, yes it does. Certainly when the EWP figures are bandied around, in order to claim wettest this, and wettest that.
In future, when we hear such claims, my recommendation is to treat them with a generous pinch of salt.
Sources
1) EWP Series
2) Met Office regional data
NOT A LOT OF PEOPLE KNOW THAT 
Decade 0 1 2 3 4 5 6 7 8 9 Decade Average
1800 157 247 218 195 313 184 258 227 127 274 220 #
1810 173 273 234 165 114 194 216 238 271 243 212 #
1820 208 148 261 251 215 219 239 171 314 179 221 #
1830 133 224 179 275 330 184 164 261 181 183 211 #
1840 270 156 207 204 166 128 242 157 306 224 206 #
1850 217 210 231 267 151 132 193 216 92 193 190 # 216
1860 243 209 154 212 164 228 280 269 230 380 237 #
1870 252 202 286 298 167 265 192 418 196 250 253 #
1880 138 268 234 331 231 264 191 241 146 171 222 #
1890 193 105 243 213 262 191 151 289 204 286 214 #
1900 322 221 225 207 283 152 246 165 222 151 219 #
1910 329 240 337 270 184 423 373 186 193 323 286 *
1920 298 209 275 332 221 330 293 199 320 170 265 * 270
1930 330 256 144 212 118 328 290 362 246 318 260 *
1940 209 264 192 292 174 232 277 235 308 193 237 #
1950 272 284 209 170 179 239 258 282 292 217 240 #
1960 374 309 257 145 88 221 363 271 211 254 250 # 246
1970 287 201 214 205 295 236 145 340 296 335 255 #
1980 326 184 215 252 317 185 272 236 293 185 247 #
1990 420 258 142 207 388 415 240 173 241 276 276 *
2000 283 331 245 277 273 164 161 327 280 219 256 * 275
2010 271 210 213 316 455 293 *
There appears to have been a major change in England and Wales rainfall, but it does not correlate with global warming. If you look carefully at the above figures:-
• Up to the end of the 19th. century, there was only ONE wet winter per decade and the average winter rainfall was 216mm (Wet winters are considered to >300mm)
• From 1910-30 the wet winters per decade suddenly increased to FOUR and the average winter rainfall increased to 270mm.
• This was way before any increase in carbon dioxide emissions. (Pre-industrial levels of carbon dioxide remain relatively constant, at 290ppm, up to 1945. They take off in the 1950’s increasing rapidly to 400ppm today)
• There was a relative dry period from 1940-80 when average winter rainfall fell to 246mm
• For the past 3 decades the wet winters seem to have returned. The average winter rainfall has increased to 275mm, similar to that experienced from 1910-30. However this CANNOT be linked to global warming.
• My analysis is not complicated.
• The Met.Office MUST know that the real “climate change” occurred in 1910, yet they are still making the claim that the recent wet winters are somehow linked to global warming!
• They should be trying to find out what caused those wetter and warmer winters to suddenly start back in 1910.
Paul, I take anything that comes out of the Met Orifice with a pinch of salt.
While talking of records, I just heard the weatherman say that this April was the third warmest since 1910. It didn’t feel like it to me.
Interesting piece Paul. The Met Office itself acknowledges the (inevitable) variation between series:
“For any individual year there can be significant differences between series”
“So which is the best series to use?
Well, the answer is that we need to use them all. The Met Office routinely quotes rainfall statistics based on the gridded data, because these are considered the most reliable estimates, are based on the full network of observations, and can provide the regional pattern of rainfall.
The EWP series is an invaluable climate series because it provides a much longer near 250-year perspective but has less regional detail.”
http://metofficenews.wordpress.com/2014/02/14/met-office-rainfall-records-how-far-do-they-go-back-and-what-can-they-tell-us/