Bob Tisdale is probably the most boring writer on WUWT. He regurgitates vast slabs of text from his own website and repeats it ad infinitum on WUWT. This time he is writing about global precipitation.
I got curious about that. Rainfall seems to me to be a tricky thing. I would expect that temperature is a lot easier to estimate. If it's hot in one place it's a fair bet it'll be hot in a neighbouring place. Rain is different. I've driven through rain bands that were no wider than 10 metres. I've sat under a downpour and then looked at the radar and seen that a storm cloud has settled over our small town and released a torrent, but that this small area is the only place that's being rained upon for hundreds of kilometres in any direction. I'm well aware that's taking a naive unscientific approach. A downpour for twenty minutes over my home town doesn't mean squat in comparison with precipitation over south eastern Australia as a whole let alone over the entire earth.
That's by way of warning - I'm not even a beginner when it comes to assessing global precipitation.
Bob Tisdale seems to be on a lower level even than I am. And I only started looking up this stuff this evening. Bob has written an article on it for WUWT, taking on the guise of an "expert".
I started running into trouble with Bob's article from his very first sentence. He writes:
We used NOAA’s CAMS-OPI satellite-era precipitation data in the post Model-Data Precipitation Comparison: CMIP5 (IPCC AR5) Model Simulations versus Satellite-Era Observations.So I looked up NOAA's CAMS-OPI and this is what NOAA says about it:
The "CAMS_OPI" (Climate Anomaly Monitoring System ("CAMS") and OLR Precipitation Index ("OPI") is a precipitation estimation technique which produces real-time monthly analyses of global precipitation. To do this, observations from raingauges ("CAMS" data) are merged with precipitation estimates from a satellite algorithm ("OPI"). The analyses are on a 2.5 x 2.5 degree latitude/longitude grid, are updated each month, and extend back to 1979. This data set is intended primarily for real-time monitoring. For research purposes, we refer users to the GPCP and CMAP products which are more quality-controlled and use both IR and microwave-based satellite estimates of precipitation.Bob Tisdale shouldn't have been using that data set to compare with model runs. It's "intended primarily for real-time monitoring".
In his latest article Bob didn't use CAMS-OPI because he wanted to look at precipitation over land separately from precipitation over the ocean. So he went to NOAA’s Global Precipitation Climatology Project (GCPC) Version 2.2. I went to see what NOAA had to say about that. Here is what they write:
One of the major goals of GPCP is to develop a more complete understanding of the spatial and temporal patterns of global precipitation. Data from over 6,000 rain gauge stations, and satellite geostationary and low-orbit infrared, passive microwave, and sounding observations have been merged to estimate monthly rainfall on a 2.5-degree global grid from 1979 to the present....
...The GPCP data have already been found capable of revealing changes in observed precipitation on seasonal to interannual time scales and in validating model generated precipitation from re-analysis systems, such as those from NCEP/NCAR and ECMWF. GPCP also offers the potential for studying changes in the distribution of precipitation at longer time scales such as predicted by GCM simulations, especially in the pattern change over previously data-sparse ocean areas. GPCP estimates can validate both the magnitude and the spatial pattern of modeled rainfall to within the estimated error of the observations. However, realization of the full potential for the GPCP to provide precipitation estimates for climate change studies, especially over the oceans, requires further research and development. Specifically, investigation of inhomogeneities in the GPCP satellite component data sets, and enhanced calibration and validation efforts, especially over open oceans, are required.How it's put together is by collecting and analysing data from rain gauges and merging it with data collected from satellites and then estimating monthly rainfall on a 2.5 degree global grid. I'm sure it's of great value for research purposes, but it's not the sort of thing I'd be comparing to models in the way I'd compare a pattern of global temperatures. A scientist skilled in this sort of analysis would be able to do this, but a blogger on HotWhopper or WUWT? No way. Here is a paper discussing GPCP, which includes a long list of other papers that cite it - a relatively small number of the 1442 cites indicated by Google.
The paper gives a solid overview but is very technical. I went looking further afield at how the data is collected. I came across a NASA website discussing TRMM, which was launched in 1997 and:
...has provided critical precipitation measurements in the tropical and subtropical regions of our planet. The Precipitation Radar (PR) can see through the precipitation column, providing new insights into tropical storm structure and intensification. The TRMM Microwave Imager (TMI) measures microwave energy emitted by the Earth and its atmosphere to quantify the water vapor, the cloud water, and the rainfall intensity in the atmosphere. TRMM precipitation measurements have made and continue to provide critical inputs to tropical cyclone forecasting, numerical weather prediction, and precipitation climatologies, among many other topics, as well as a wide array of societal applications.
Then there is the Global Precipitation Measurement (GPM) mission , a joint project initiated by NASA and JAXA. The core GPM observatory is scheduled for launch next year. You can read about it here.
To see more about how accurate is the measured precipitation I turned to Google. Here is a 2010 paper discussing uncertainties in satellite precipitation data sets in the TRMM era. The abstract states in part:
The uncertainties are relatively small (40–60%) over the oceans, especially in the tropics, and over southern South America. There are large uncertainties (100–140%) over high latitudes (poleward of 40° latitude), especially during the cold season. High relative uncertainties are also evident through the seasons over complex terrain areas, including the Tibetan Plateau, the Rockies and the Andes. Coastlines and water bodies also indicate high measurement uncertainty. The estimated global uncertainties also exhibit systematic seasonal, regional as well as rain-rate dependencies, with lowest uncertainties over tropical oceanic regions with strong, convective precipitation, and highest ones over wintery, complex land surfaces with light precipitation.
There is much more information out there, including lots of recent papers. I'm intrigued by how much scientists can tell us about precipitation, by clever use of technology. See, WUWT can't be all bad :)
As far as I can see, Bob Tisdale is not proving or disproving anything in his crude comparison of monthly global precipitation with his multi-model means from CMIP5. He takes the observations at face value for one thing, without allowing for uncertainties in either the observations or the models.
And that, of course, is apart from the fact that just as in the past, unless his ensemble mean lines up exactly with observations, Bob is inclined to dismiss the models altogether.
It is clear that these people are only interested in statistical significance when it suits them. Apparently, it does not suit them when it comes to comparisons between models and observations.
ReplyDeleteIf they (hypothetically) were interested, they would have to look at the variability in the individual model runs and not in their average (the latter is by mathematical necessity much smaller).
The bigger issue with precipitation for every day life and infrastructure planning, is how heavy it is, when it occurs and where it occurs. Are there going to be seasonal shifts, which will affect agriculture and water storage planning. What about floods - like Toronto and Calgary recently etc.
ReplyDeleteGlobal precipitation trends don't have all that much relevance compared to regional patterns in my view. (Unless it was to change markedly all over I suppose.)