Hi Mike! Yep, there's never been consistency in the literature with respect to adjectives for spatial resolution, probably because of the fact that these are relative terms and different communities have different baselines and the baselines change with technology. So, AVHRR was "Very High Resolution" for the meteorological community but was always low resolution for the terrestrial ecology community. Alan Strahler's classic paper (Strahler et al. (1986) On the nature of models in remote sensing. Remote Sensing of Environment 20: 121-139) is very useful in thinking about this relativism. In response to Sharon, I would say the term "scale" is much less consistently used in the literature (see Dungan et al. (2003) A balanced view of scale in spatial statistical analysis. Ecography 25, 626-640) because this term can refer to several different aspects of the data, the process, and the models used to analyze the data and process. I think your nomenclature works well. It differs from some others (e.g. http://www.wmo-sat.info/oscar/instrumenttypes and http://www.eohandbook.com/eohb2014/sat_earth_obs_high_res_optical.html) but seems more logical.

if you speak in terms of satellite image maps I think it is much more productive to adopt a classification relating spatial resolution to standard map scales. like 1Millon 500k 50k etc. But take in account that the value of classifactions is relative, the best classification for a work strongly depends on the purpose of your study and if is it usefull or not. P.e. MODIS ihas a perfect spatial resolution for cloud and regional studies, Ikonos has very high spatial resolution but is not usefull at all for cloud studies.

I rather like Moeller's classification (see attached). But I do agree that there is no standardized descriptions for geometric resolution from aerial and spaceborne platforms. I think the greatest hurdle in this area has been technology itself continuing to improve resolution and so a definition (if one exists) continues to be redefined. If you do have any articles along this area I'd like to see them if you don't mine.

I agree that there is a lack in consistency in how spatial resolution ranges are described. Two factors that possibly contribute are field of application and historical terminology.

If the application is at a global scale (e.g. vegetation monitoring) then a spatial resolution of say 30m is high. On the other hand, if the application is at a very local scale (e.g. urban studies) then a spatial resolution of 30m is low.

The rapid change in technology, as Cristina and Ron mentioned, has changed how we in general perceive spatial resolution. Sensors such as SPOT's HRVIR (high resolution visible and infrared) were at the time of their launch considered high resolution with multispectral resolution of 20m, while these days it would probably be considered medium (this is of course open to debate - see my first point).

So I think there will always be inconsistency across the board in how spatial resolution is defined but I do think that within application fields there is usually some consistency. I agree with Cristina and Gabriel that perhaps the use of scale is clearer and that we should always be specific about what we are using or refering to.

Hi Mike! Yep, there's never been consistency in the literature with respect to adjectives for spatial resolution, probably because of the fact that these are relative terms and different communities have different baselines and the baselines change with technology. So, AVHRR was "Very High Resolution" for the meteorological community but was always low resolution for the terrestrial ecology community. Alan Strahler's classic paper (Strahler et al. (1986) On the nature of models in remote sensing. Remote Sensing of Environment 20: 121-139) is very useful in thinking about this relativism. In response to Sharon, I would say the term "scale" is much less consistently used in the literature (see Dungan et al. (2003) A balanced view of scale in spatial statistical analysis. Ecography 25, 626-640) because this term can refer to several different aspects of the data, the process, and the models used to analyze the data and process. I think your nomenclature works well. It differs from some others (e.g. http://www.wmo-sat.info/oscar/instrumenttypes and http://www.eohandbook.com/eohb2014/sat_earth_obs_high_res_optical.html) but seems more logical.

I think the spatial resolution of RS data is relevant to the objectives of the application. Depending upon the extent and type of information to be extracted, the resolution needs to be defined. I agree that scale is the better and standard way of defining the SR.

I don't get hung up on VHSR to VLSR. The most important thing I look for is the definition for whatever the authors thinks is high or low within their spatial neighborhood. With ground-based active-optical sensors, VHSR as defined by you in your question isn't very high at all. Everything to relative scale. Just look for the definition.

I agree with Cristina's comment that cartographic scale and spatial resolution are not the same. In the case of RS images, and assuming that the pixel size coincides with the ground-projected instantaneous field of view (GIFOV) of the sensor, the relation depends on the dots per inch (DPI) or pixels per inch (PPI) of the plotter/display. Different monitors have different PPIs, ranging from 300, so there is no one-to-one relationship.

In any case, I like the intervals you propose, Mike, because they follow orders of magnitude in the metric system. I just would change the name of the intervals, so that they are application-independent (abounding on Jennifer's comment , HR vs LR depends on the size of the objects of interest, sensu Woodcock and Strahler 1987):

< 1 m - submetric spatial resolution (sSR, i.e., less than a meter);

1-10 m - metric spatial resolution (mSR, i.e., in the order of meters);

10-100 m - decametric spatial resolution (dSR, i.e., in the order of decameters);

100-1000 m - hectometric spatial resolution (hSR, i.e., in the order of hectometers);

> 1000 m - kilometric spatial resolution (kSR, i.e., in the order of kilometers).

Actually, the 'submetric' apellative is already common. So I don't see why the remaining interval names could not catch up if folks started to use this...

Thank you to everyone for the insightful replies. Very much appreciated!

I'd like to add a further comment that we should always use the word "spatial" in this regard to clarify what is being discussed. As there are now super / hyper spectral systems and the capabilities to acquire multi-temporal datasets increases, the "important" resolution may not be restricted to the spatial domain.

At a recent meeting in Berlin the rather confusing use of the terms high and medium spatial resolution for 20 - 30 m systems was apparent between the European and North American delegates.

Cheers,

Geoff

Hi. I agree with Geoff. I think the key of the discussion here it is that there are many kind of "resolution", for example for the mentioned AVHRR, "high resolution" is related to radiometry (10 bits) not to range and it was an improvement in comparison with older meteorologic sensors.

Hello Michael and others,  I am a relatively new comer to research gate and just came across this year old question and responses about how adjectives used for spatial resolution can cause confusion.  A lot of good comments already posted and i agree with what they have to say.  Like the saying 'one person's trash is another person's treasure' .......... this kind of applies to what resolution is 'high' and what is 'low'.  As has been mentioned it is more or less based on the application of interest.  The suggestion to right away give the spatial resolution being used would be best; I would add avoid using the adjectives of low, moderate, high or what ever else might be used to descibe the data from that particular user's point of veiw.

As mentioned now days 'resolution' does not apply only to 'spatial' but also temporal, and I would add 'spectral'.  To some people 50 spectral bands is 'hyper' but to others 256 bands is 'hyper'; so for me it would be best to just give the spatial, temporal, spectral, and radiometric (8 bits vs 11 bits vs xx bits)  characteristics of the image data being used in the analyses and let the reader decide what adjective he/she might want to use.

Sorry for the somewhat 'late' response.

Pat