Guest Post by Willis Eschenbach
Anthony has posted on a recent study (behind a paywall) of rice production in Rice yields, CO2 and temperature – you write the article. The article claims that rice yields are falling, and will fall further, due to temperature changes.
I said I’d write the article if someone would send me the actual study, and a couple of WUWT readers came through, my thanks to them. Here’s my take on what the authors have done.
First, the good. They have used actual data from farmer’s fields, rather than theoretical data or greenhouse experiments.
Figure 1. Rice Fields in Ha Giang
However, there are some troubling things in the study.
First, it covers a very short time span. The longest farm yield datasets used are only six years long (1994-99). Almost a fifth of the datasets are three years or less, and the Chinese data (6% of the total data) only cover two years (1998-1999).
Now, if they were comparing the datasets to temperature records for the area where the farms are located, we could get useful information from even a two-year dataset. But they are not doing that. Instead, they say:
Data series from the weather stations at the sites were too short to determine trends. Instead, trends in Tmin and Tmax were based on a global analysis of ground-station data for 1979–2004 …
Unfortunately, they have neglected to say which “global analysis of ground-station data” they are using.
But whichever dataset they used, they are comparing a two year series of yields against a twenty-six year trend. I’m sorry, but I don’t care what the results of that comparison might be. There is no way to compare a two-year dataset with anything but the temperature records from that area for those two years. This is especially true given the known problems with the ground-station data. And it is doubly true when one of the two years (1998) is a year with a large El Niño.
For example, they give the trend for maximum temps in the winter (DecJanFeb) for the particular location in China (29.5N, 119.47E) as being 0.06°C per year, and the trend for spring (MarAprMay) as being 0.05°C per year (I get 0.05°/yr and 0.04°C/yr respectively, fairly close).
But from 1998 to 1999, the actual DJF change was +2.0°C, and the MAM change was minus 1.0°C (CRU TS Max Temperature dataset). As a result, they are comparing the Chinese results to a theoretical trend which has absolutely no relationship to what actually occurred on the ground.
To try to find out which “ground-station data” they used, I compared their temperatures for China (29.5N, 119.47E) with the CRUTEM3, CRU TS, and GISS records from KNMI. However, I could not match their numbers, although I could get sorta close. Worrisome.
Next, they have not mentioned autocorrelation anywhere in their study. This makes me think that they have not adjusted for autocorrelation, which is particularly important with short datasets.
Next, they base their predictions for the future on a single computer model of the regional changes. It is widely agreed that computer climate models are not very good at predicting regional changes, so that part of their study seems very weak.
I am most mystified by their use of the 26-year temperature trend. Why not use the actual year-by-year changes in the local temperature? The rice is responding to actual temperatures, not to a mathematical trend … so why not compare yields to actual temperatures?
So once again, we have questionable methods used with uncited data to give alarming results. It is too bad, because their premise is good, and so is their general approach (use actual farm data).
Color me unimpressed …